JP2000054862A - Output control method in vehicle with power source - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reflect the taste or ability of a user in the output performance of a vehicle with power source by evolving the relation between the operation quantity of a primary output regulating means and the operation quantity of a secondary output regulating means by use of genetic algorithm on the basis of either one of characteristics of the user, the using state, and the change of the using environment. SOLUTION: User evaluation information is inputted to the development evolution control part of a control unit through a push-button type evaluation value input value, and the evolution control part evolves the relation between input information and output information, or the relation of the current carrying signal to the evolution control part for controlling the drive quantity of a stepping motor for driving a throttle valve as a secondary output regulating means or controlling the current carrying to an electric motor to the operation quantity by the user of an accelerator grip as a primary output regulating means by use of genetic algorithm along the evaluation of the user.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an improvement in an output control method for a vehicle with a power source using an engine or a motor as a power source.

[0002]

2. Description of the Related Art A vehicle with a power source using an engine or a motor as a power source is provided with a primary output adjusting means such as an accelerator pedal, an accelerator lever or an accelerator grip so that a user can adjust the output of the power source. Is provided,
With this primary output adjusting means, secondary output adjusting means such as a throttle valve for directly adjusting the output of the power source and a current supply control unit for the motor are operated, and finally the output of the power source is adjusted. . And an operation amount of the primary output adjustment means,
The operation amount of the secondary output adjustment means is determined and fixed in a predetermined relationship at the time of shipment, and therefore, for example, when the accelerator grip is adjusted to a certain rotation angle, the throttle valve basically always has the same opening degree at all times. In addition, the amount of fluctuation of the throttle valve according to the amount of fluctuation of the accelerator lever is always constant.

[0003]

The relationship between the operation amounts of the primary output adjusting means and the secondary output adjusting means described above is directly expressed as the output performance of a vehicle with a power source. On the other hand, the larger the throttle opening is, the larger the output is, so the power is emphasized.The smaller the throttle opening is, the smaller the output is, the fuel consumption is emphasized, and the variation of the throttle valve with respect to the variation of the accelerator grip is reduced. A larger one is a response-oriented type, and a smaller throttle valve fluctuation amount is a smaller output fluctuation. Since the preference for the output performance of the vehicle with the power source described above differs for each user, it is assumed that the relationship between the primary output adjustment means and the secondary output adjustment means is determined to be a predetermined relationship at the time of shipment as in the related art and is fixed. However, there is a problem that preferences of individual users are not reflected. In addition to the above-mentioned problems, vehicles with a power source are used in various environments, and their use conditions vary from user to user. Furthermore, the power source (including peripheral members) deteriorates with use. Therefore, there is also a problem that the relation between the primary output adjustment means and the secondary output adjustment means set at the time of shipment cannot always be maintained in an optimum relation. SUMMARY OF THE INVENTION The present invention solves the above-mentioned conventional problems, and solves the characteristics (preference, skill, state) of each user, the running condition of a vehicle with a power source, changes in a driving environment, and the aging of a power source or its peripheral members. It is an object of the present invention to provide an output control method for a vehicle with a power source that can adaptively change the relationship between the primary output adjustment means and the secondary output adjustment means.

[0004]

SUMMARY OF THE INVENTION In order to achieve the above object, an output control method for a vehicle with a power source according to the present invention is provided. An output control method of a power source mounted on a vehicle with a power source for determining an operation amount of a secondary output adjustment means for adjusting an output of a power source, wherein an operation amount of the primary output adjustment means and a secondary output adjustment means A genetic algorithm based on at least one of a user's characteristics, a use state, a change in a use environment, and a temporal deterioration of a power source. .

[0005]

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a perspective view of a vehicle equipped with a power source according to an embodiment of the present invention; FIG.
FIG. 12 to FIG. 12 show an embodiment in which the output control method of a vehicle with a power source according to the present invention (hereinafter, simply referred to as an output control method) is applied to an internal combustion engine equipped with an electronic throttle. FIG. 1 is a schematic diagram showing a relationship between an engine and a control unit that executes an output control method. The engine shown in the drawing detects the operation amount of the accelerator grip by the user with an accelerator operation amount detection sensor, drives a stepping motor based on the detected accelerator operation amount, and adjusts the opening of the throttle valve with the stepping motor. Equipped with a so-called electronic throttle, the control unit
Inputting a crank angle signal and an accelerator operation amount signal obtained from a crank angle sensor provided on the engine and an accelerator operation amount detection sensor provided on the accelerator grip,
The drive signal of the stepping motor is determined and output based on the input information. FIG. 2 is a schematic block diagram showing the internal configuration of the control unit. As shown in FIG. 2, the control unit includes an engine speed calculation unit and an evolution control unit, and the engine speed calculation unit calculates the engine speed based on the crank angle signal. The evolution control unit inputs the engine speed and the accelerator operation amount signal calculated by the engine speed calculation unit, and determines and outputs a drive signal of the stepping motor based on the input information. As shown in FIGS. 1 and 2, the user evaluation information is input to the evolution control unit of the control unit via a push button type evaluation value input device. The evolution of the relationship between the input information and the output information, that is, the drive characteristic of the throttle valve with respect to the amount of accelerator operation by the user, allows the output characteristic of the vehicle equipped with the engine to be changed by the user. It is configured so that it can be adapted to characteristics that match the taste of the user. Here, the characteristics of the electronic throttle valve with respect to the accelerator operation amount will be briefly described. The drive characteristics of the electronic throttle valve with respect to the accelerator operation amount include static characteristics and dynamic characteristics. The static characteristic is a relationship between the accelerator operation amount and the throttle valve opening when the accelerator operation amount (that is, the operation rotation angle of the accelerator grip) is constant, and affects the steady running characteristics of the vehicle. By changing the static characteristics in this manner, as shown in FIG. 3, the electronic throttle valve opens greatly when the operation rotation angle of the accelerator grip is small, and the throttle valve gradually opens fully as the operation rotation angle of the accelerator grip increases. Converging low opening rapid acceleration type (Fig. 3
(See FIG. 3A), a proportional type in which the operation angle of the accelerator grip is proportional to the throttle opening (see FIG. 3B),
Alternatively, the throttle valve is gradually opened when the operation rotation angle of the accelerator grip is small, and is rapidly opened to a full opening when the operation rotation angle of the accelerator grip is large.
(See (c)), various opening degrees of the throttle valve can be obtained at the same operation rotation angle of the accelerator grip. This static characteristic is defined as any function as long as the opening of the throttle valve increases or does not change with an increase in the accelerator operation amount, and the opening of the throttle valve is zero when the accelerator operation amount is zero. By changing the static characteristics, it is possible to output different throttle valve openings for the same accelerator operation amount. The dynamic characteristic is a relationship between a change in the opening degree of the throttle valve and a change in the accelerator operation amount, and affects a transient characteristic of the vehicle. Specifically, it can be changed by combining a first-order lag filter and an incomplete differential filter and changing these parameters. By changing the dynamic characteristics in this way,
As shown in FIG. 4, the throttle responds relatively slowly to the accelerator operation, and the throttle opens slowly (see FIG. 4).
(A), a slight spike occurs with respect to the accelerator operation, but the type changes quickly and the throttle opens (see FIG. 4 (b)), or a type intermediate between the two (see FIG. 4 (c)). Various dynamic characteristics can be obtained.

Next, the configuration of the evolution control unit will be described in detail. FIG. 5 is a schematic block diagram of the evolution control unit. As shown in the drawing, the evolution control unit includes a reflection layer, a learning layer, and an evolution adaptation layer.

(Reflection Layer) The reflection layer comprises an electronic throttle control module and an electronic throttle control parameter initial value setting section. As shown in FIG. 6, the electronic throttle control module includes a static characteristic changing unit and a dynamic characteristic changing unit including a first-order lag filter and an incomplete differential filter, and an actual accelerator operation amount (accelerator operation amount signal). x1 is input, and the actual accelerator operation amount x1 is set by the static characteristic changing unit. The virtual accelerator input x according to the static characteristic set.
The dynamic characteristic changing unit determines the opening of the electronic throttle valve from the virtual accelerator input x2 and outputs it. FIG.
Where x1 is the actual accelerator operation amount (accelerator operation amount signal), x2 is the virtual accelerator input, y is the opening of the accelerator valve, f is the static characteristic function, T is the first-order lag time constant, Td is the differential time, and α is The acceleration correction coefficient and η indicate the differential gains. Among them, the static characteristic function, the first-order lag time constant, the acceleration correction coefficient, the differential time or the differential gain are controllable control parameters. In the example, it is assumed that the user selects a static characteristic function from several prepared ones in advance, the differential gain of the dynamic characteristic is fixed, and the primary delay time constant, the acceleration correction coefficient, and the differential time are evolved. explain. In the following description, the first-order lag time constant, the acceleration correction coefficient, and the derivative time are collectively referred to as electronic throttle control parameters. The electronic throttle control parameter initial value setting unit determines and outputs initial values of the above-described first-order lag time constant T, acceleration correction coefficient α, and derivative time Td.

(Evolution Adaptation Layer) In the evolution adaptation layer,
The correction values of the electronic throttle control parameters are evolved so that the electronic throttle control parameters in the reflective layer are evolved adaptively to the user's preference. As shown in FIG. 5, the evolution adaptation layer includes an evaluation unit and an evolution adaptation unit. The evolution adaptation unit has a control module that determines the electronic throttle control parameter correction value in the reflective layer.The drivability evaluation unit uses the evaluation of each individual when evolving the control module of the evolution adaptation unit using a genetic algorithm. Perform based on the evaluation of the person. FIG. 7 is a diagram showing a specific example of a control module in the evolution adaptive layer. As shown in FIG. 7, the control module uses the normalized accelerator operation amount and the normalized engine speed as input information, and Electronic throttle control parameters, ie, first order lag time constant T, acceleration correction coefficient α, and correction amount A of differential time Td
It consists of a two-input, three-output neural network using T, Aα, ATd as output information. In the evolution adaptation layer, several chromosomes (solids) are generated by coding the connection weight of the neural network constituting each control module of the evolution adaptation unit as a gene, and each generated individual is selected according to the evaluation of the evaluation unit. At the same time, the remaining solids are crossed to generate next-generation solids, and the selection of these is repeated, whereby each control module of the evolution adaptation unit is evolved in accordance with the evaluation of the evaluation unit.

The above-described evolution processing will be described more specifically. FIG. 8 is a flowchart showing the flow of the control module evolution process using the genetic algorithm. First, as shown in FIG. 9, a plurality of individuals a (n) (n = 1 in this embodiment) are coded by using the coupling coefficient of the neural network constituting the control module as a gene.
0) is generated (step 1). Here, the initial value of the gene value of each individual (that is, the value of the coupling coefficient of the neural network) is within a predetermined range (approximately −
(Between 10 and 10). At this time,
If the learning layer has already performed learning and output, by including one individual that can make the output of the evolutionary adaptive layer zero, without impairing the performance at that point even if the number of individuals is limited The diversity of the population during the evolution process can be maintained. Next, the connection weight of the neural network of the control module is fixed with the connection weight of one of the individuals a (n) generated in Step 1, for example, the connection weight of the individual a (1), and the actual input information ( The outputs AT, Aα, and ATd of the neural network with respect to the engine speed and the accelerator operation amount are determined (step 2), and the outputs are linearly transformed using the equations (1-3) to obtain an individual an (1). The outputs ATy1, Aαy1, and ATdy1 of the control module (that is, electronic throttle valve control parameter correction amounts) are determined (step 3). It should be noted that the engine speed and the accelerator operation amount of the input information are each normalized. ATy1 = 2 × GAT-G (1) Aαy1 = 2 × Gα-G (2) ATdy1 = 2 × GATd-G (3) where ATy1, Aαy1, ATdy1 control the output AT, Aα, ATd of the control module. The output of the neural network in the module, G, is the evolution adaptive layer output gain. In this manner, by using the output of the neural network after linear conversion, the output AT from the control module is obtained.
y1, Aαy1, and ATdy1 do not become extremely large values, and the evolution proceeds slightly as a whole, and the behavior of the engine does not extremely fluctuate due to evaluation and evolution. After determining the output ATy1, Aαy1, ATdy1 of the control module for the individual a (1), the output AT of the learning layer for each control parameter correction value
y2, Aαy2, ATdy2 are determined (step 4),
The vehicle is actually run with the electronic throttle control parameters in the reflective layer corrected by the correction amounts Y1, Y2, and Y3 obtained by adding the output of the control module and the output of the learning layer (step 5). The evaluation of 1) is input (step 6), and the evaluation value of the fitness of the individual a (1) is determined (step 7). The output of the learning layer is set to zero in the initial state, and the learning layer may be configured to learn the result of the evolution after the evolution process in the evolution adaptive layer is completed. A detailed description of the learning layer will be described later. The evaluation of the individual by the user in step 6 described above may be input through an evaluation value input device configured as a push button so as to be operated during driving. Specifically, for example, when the driver presses this button, the evaluation value of the individual is determined in step 7 based on the length of time the button is pressed. The evaluation value calculation method in step 7 includes, for example, a method of multiplying the reciprocal of the pressed time by a constant coefficient, and a method of calculating from the length of the pressed time using a fuzzy rule. In this way, the evaluation value can be obtained with a certain degree of accuracy even if there is ambiguity in human evaluation. Furthermore, if the user keeps pressing this button for more than a certain period of time, the individual being evaluated at that time can be eliminated. By doing so, the driver can immediately eliminate individuals having unfavorable characteristics, and since this individual does not affect the next generation, evolution can be performed at high speed. The processing of steps 6 and 7 is performed by the evaluation unit. The processing from step 2 to step 7 is performed on all the individuals generated in step 1, and if it is determined that the fitness of all the individuals has been evaluated (step 8), the predetermined generation number evolution processing is performed. It is determined whether or not it has been performed (step 9). If the evolution process has not been performed for a predetermined number of generations, a next-generation parent individual is selected from the ten individuals generated in step 1 (step 10). For this selection, a roulette-type selection method is used, and some parent individuals are selected stochastically with a probability proportional to the fitness of each individual. At this time, if the generational change is applied too strictly, individuals with high evaluations may be destroyed. Therefore, an elite preservation strategy that unconditionally leaves the elite (the individual with the highest evaluation) to the next generation is also combined. Used. In addition, the fitness is linearly converted so that the ratio between the maximum fitness and the average fitness of a plurality of individuals is constant. When the selection of the parent individual is completed, crossover is performed using the selected individual as a parent individual to generate a second generation of ten child individuals again (step 11). For crossover between individuals, a technique such as one-point crossover, two-point crossover, or normal distribution crossover is used. The normal distribution crossover is a method of generating children according to a normal distribution that is rotationally symmetric with respect to an axis connecting parents with respect to a chromosome (individual) represented by a real value. The standard deviation of the normal distribution is proportional to the distance between the parents for the main axis component connecting the parents, and proportional to the distance between the straight line connecting the parents and the third parent sampled from the population for the other axis components. . This crossover method has the advantage that the characteristics of the parent are easily inherited by the child. Also, the gene (coupling degree) value is randomly changed with a certain probability for the generated 10 offspring individuals, thereby causing gene mutation. Note that these ten offspring individuals include one individual that can make the output of the evolutionary adaptive layer zero. After the second generation is generated by the above processing, the evolution processing from step 2 is repeated again. The above-described evolution processing is repeatedly performed until a predetermined number of generations has elapsed. Thereby, the child individuals constituting each generation are selected by the evaluation unit, that is, based on the preference of the user, and the next generation is generated. Therefore, the input / output relationship of the first control module is as follows. It gradually evolves to suit the taste of the user. Whether or not a predetermined number of generations has elapsed is determined in step 9, and if it is determined in step 9 that it is the last generation, an individual with the highest fitness (optimal individual) among the 10 child individuals of that generation That is, one elite is selected (step 12), the coupling coefficient of the neural network of the control module is fixed with the gene constituting the above-mentioned optimal individual (step 13), and the learning process for the learning control module of the learning layer is performed. Move to

(Learning Layer) Next, the learning layer will be described in detail. The learning layer is the output of the control module after evolution obtained in the evolution adaptation layer (in the case of this embodiment,
The electronic throttle control parameter correction amount is learned, and the output of the evolution adaptive layer is reflected on the reflection layer even after the evolution processing is completed.
If the output of the control module of the evolution adaptive layer is information that does not depend on the driving state of the vehicle, that is, if there is no information on the driving state in the input of the control module, the output simply evolves to the output of the control module of the learning layer. It is only necessary to add the value obtained by adding the outputs of the control modules of the adaptive layer to the output of the learning layer, but if the output of the control module of the evolution adaptive layer is information that depends on the driving state of the vehicle, the learning layer Since it is necessary to learn the relationship between the operation state of the vehicle and the output of the control module of the evolution adaptive layer, the learning layer is provided with a control module for learning and a control module for execution.
In the case of the present embodiment, since the control module of the evolution adaptive layer is configured by a neural network that uses the engine speed, which is one of the driving states of the vehicle, as input information, the learning layer adopts the latter configuration. . Therefore, the learning layer includes a control module corresponding to the control module of the evolution adaptation layer, and this control module includes two control modules (not shown), and one of the two control modules functions as a control execution. During this operation, the other control module functions as a learning module so that the functions can be interchanged with each other. The control module may be any control module as long as it can be learned. For example,
Neural network, CMAC and the like. In the learning layer, when the evolution processing of each module is performed by the evolution adaptation layer for a predetermined number of generations, and the neural network constituting the module is fixed with the optimal coupling coefficient at that time, the input of the control module of the evolution adaptation layer is performed. The teacher data is acquired by matching the relationship between the input and the output with the relationship between the input and the output of the control module functioning for executing the learning layer. In particular,
As shown in FIG. 10, the input and output of the evolution adaptive layer and the learning layer are averaged with a certain step width, and this is used as teacher data. For example, the average engine speed per second is 5000 rpm,
When the average opening degree is 20%, the sum of these and the electronic throttle control parameter correction amount of the evolution adaptive layer and the learning layer at this time is used as teacher data. Also,
At this time, the obtained creation time of the teacher data is recorded, and when learning is performed in the learning layer, the newer creation time is regarded as being more important as teacher data. Upon acquiring the teacher data of the control module, the learning control module of the learning layer starts learning based on the acquired teacher data, and outputs the outputs of the learning control module and the execution control modules in the evolution adaptive layer and the learning layer. The learning is terminated when the difference from the threshold value becomes smaller than the threshold value. After the learning ends, the control module is replaced with the control module for execution, and the output of each control module of the evolution adaptive layer is set to zero or the next evolution processing output. During the learning by the learning control module, the control module of the evolution adaptive layer keeps outputting the electronic throttle control parameter correction amount with the optimal coupling coefficient fixed, and outputs the evolution adaptive layer and executes the learning layer. The electronic throttle control parameter corrected by the correction amount obtained by adding the output of the electronic control module is used for the calculation by the electronic throttle control module in the reflection layer. The initial value of the execution control module in the learning layer is set so that the output is always zero. By doing so, in the initial state, control in the reflection layer can be performed using only the output of the evolution adaptive layer. The information of the control module learned in the learning layer can be stored in the internal storage means, the external storage means, or the like, and the user can call the stored characteristics as necessary and travel with the characteristics. In this way, by allowing the learning result to be stored and recalled, it is possible to cope with a change in the mental state of the user. When the control module in the learning layer is a neural network, learning is performed by a normal learning method. However, when the control module is configured by CMAC, a new acquisition is performed as shown in FIG. Only the part corresponding to the learned teacher data can be learned, and the learning efficiency is improved.

Here, the flow of the entire evolution process in the evolution control unit described above will be briefly described. As shown in FIG. 12, when the evolution process is started, the evolution control unit firstly sets the throttle static characteristic. Settings are made (step 1). The setting of the static characteristic may be selected and set by the user from several static characteristics prepared in advance, or may be set to a specific static characteristic in advance. When the setting of the static characteristics is completed, the evolution adaptive layer evolves the control module while repeating generation, evaluation, and selection of the individual as described above,
After reaching the predetermined number of generations, the learning layer is made to learn the evolution result (step 2). When the learning is completed in the learning layer, the evolution adaptive layer stops the output, and the electronic throttle control by the reflection layer and the learning layer, that is, the electronic throttle control using the electronic throttle control parameter corrected by the correction value output from the learning layer. Control is performed (step 3). After that, the evolution adaptation layer is activated at regular time intervals, the drivability performance is evaluated.If the performance does not improve, the control by the reflection layer and the learning layer is continued. The evolution process is automatically started from the process 2 (step 4). Note that the evolution process may be configured to start immediately after the engine is started, or may be configured to be started by a user's start instruction. Also, as in the present embodiment, after the evolution processing is completed once, the evolution adaptation layer may be activated at regular time intervals to start the evolution processing again when the performance is improved. May be configured to start the evolution process in response to the instruction.

Next, an embodiment in which the output control method according to the present invention is applied to electric motor drive control in an electric motor drive type vehicle will be described. FIG. 13 is a schematic diagram showing the relationship between the electric motor and a control unit that executes the output control method. The electric motor shown in the drawing is a three-phase AC servomotor, and the control unit detects an operation amount of an accelerator grip (not shown) by a user with an accelerator operation amount detection sensor, and based on the detected accelerator operation amount. To determine the current flowing from the battery to the motor. FIG. 14 is a schematic block diagram showing the internal configuration of the control unit. As shown in FIG. 14, the control unit calculates the motor rotation speed and the magnetic pole position based on the encoder signal, and calculates the motor rotation speed and the magnetic pole position, based on the accelerator operation amount detected by the accelerator operation amount detection sensor. Motor current command value calculation unit (evolution control unit) that determines the current command value to the motor, and converts the DC current from the battery into an AC current based on the motor current command value obtained from the motor current command value calculation unit. And a motor energization control unit for energizing the electric motor with alternating currents having different phases. FIG. 15 is a block diagram illustrating a schematic configuration of the motor power supply control unit. As shown in the drawing, the motor power supply control unit includes a motor current command value from a motor current command value calculation unit (evolution control unit). A current control calculation unit that performs current control on the motor based on the fed-back motor currents 1 and 2;
PW based on the control signal output from the current control calculation unit
It has a PWM waveform generator for generating an M waveform, and a gate drive circuit. As shown in FIG. 13 and FIG. 14, user evaluation information is input to a motor current command value calculation unit (evolution control unit) of the control unit via a push button type evaluation value input device. The command value calculation unit (evolution control unit) evolves the relationship between the input information and the output information, that is, the current command value to the electric motor with respect to the accelerator operation amount of the user, based on the user evaluation information. By changing, the output of the vehicle on which the engine is mounted can be adapted to the drivability characteristics suitable for the user's preference. Here, the characteristics of the current command value to the electric motor with respect to the accelerator operation amount will be briefly described. The characteristics of the current supplied to the electric motor with respect to the accelerator operation amount include static characteristics and dynamic characteristics. The static characteristic is a relationship between the accelerator operation amount and the current command value when the accelerator operation amount (that is, the operation rotation angle of the accelerator grip) is constant, and affects the steady running characteristics of the vehicle. By changing the static characteristics, as shown in FIG. 16, the energizing current when the operation rotation angle of the accelerator grip is small is relatively large, and the energization current converges to the maximum as the operation rotation angle of the accelerator grip increases. Low opening rapid acceleration type (Fig. 1
6 (a)), a proportional type in which the operating current of the accelerator grip is proportional to the energizing current (see FIG. 16 (b)), or the energizing current is gradually increased while the operating angle of the accelerator grip is small. Therefore, various energizing currents are obtained at the same operating angle of the accelerator grip, such as a high opening rapid acceleration type (see FIG. 16 (c)) in which the energizing current suddenly increases to a maximum as the operating angle of the accelerator grip increases. Will be able to do it. This static characteristic may be any function as long as the conduction current increases or does not change with the increase in the accelerator operation amount, and the conduction current is zero when the accelerator operation amount is zero. , It becomes possible to output different energizing currents for the same accelerator operation amount. The dynamic characteristic is a relationship between a change in the current command value and a change in the accelerator operation amount, and affects a transient characteristic of the vehicle. Specifically, it can be changed by combining a first-order lag filter and an incomplete differential filter and changing these parameters. FIG. 17 shows some dynamic characteristics when the electric motor is driven, and FIG.
Shows some dynamic characteristics during regeneration of the electric motor.

Next, the configuration of the evolution control unit will be described in detail. FIG. 19 is a schematic block diagram of the evolution control unit. As shown in the drawing, the evolution control unit includes a reflection layer, a learning layer, and an evolution adaptation layer, as in the first embodiment.

The reflection layer comprises a motor control module and a motor control parameter initial value setting section. The motor control module includes a static characteristic changing unit and a dynamic characteristic changing unit including a first-order lag filter and an incomplete differential filter, as in the first embodiment. The changing unit converts the actual accelerator operation amount into a virtual accelerator input according to the set static characteristic, and the dynamic characteristic changing unit determines and outputs a conduction current from the virtual accelerator input. The initial value of the control parameter used in the static characteristic changing unit and / or the dynamic characteristic changing unit in the motor control module is set by the motor control parameter initial value setting unit.

The evolution adaptation layer evolves the correction values of the motor control parameters so that the motor control parameters in the reflection layer evolve adaptively to the user's preference. The evolution adaptation layer includes an evaluation unit and an evolution adaptation unit. The evolution adaptation unit includes a control module that determines a motor control parameter correction value in the reflection layer. The drive evaluation unit evaluates each individual when evolving the control module of the evolution adaptation unit using a genetic algorithm. Perform based on the evaluation of The control module in the evolution adaptive layer, specifically, any of the normalized accelerator operation amount, the normalized motor rotation speed, and the normalized magnetic pole position as input information,
It comprises a neural network that uses a correction amount of a motor control parameter (for example, a first-order lag time constant, an acceleration correction coefficient, a differential time, or a static characteristic function) in the reflection layer as output information. In the evolution adaptation layer, several chromosomes (solids) are generated by coding the connection weight of the neural network constituting each control module of the evolution adaptation unit as a gene, and each generated solid is subjected to a user's evaluation. The control module of the evolution adaptation unit is evolved in accordance with the evaluation of the evaluation unit by repeating the selection, the generation of the next-generation solids by crossing the remaining solids, and the elimination of these. Since the evolution processing using the genetic algorithm in the evolution adaptation layer is basically the same as the evolution processing of the first embodiment, a detailed description is omitted here.

As in the first embodiment, the learning layer includes a learning neuro and an execution neuro, and while the control is being executed by the execution neuro, the control module whose evolution processing has been completed by the evolution adaptation layer. The information is learned by the learning neuro, and when the learning is completed, the learning neuro and the execution neuro are switched.

As in the first and second embodiments described above, the operation amount of the accelerator grip as the primary output adjustment means,
The relationship between the drive amount of the stepping motor that drives the throttle valve as the secondary output adjusting means or the energizing current signal to the control unit that controls the energizing of the electric motor, using a genetic algorithm along with the user's evaluation By evolving, it is possible to obtain output performance that suits individual users' preferences, and since the user can select evolving individuals by himself, training the vehicle as if to the user It is possible to give fun as if you were.

In the embodiment described above, the user sets in advance the static characteristics of the secondary output adjusting means (stepping motor or power supply control unit) with respect to the primary output adjusting means (accelerator operation amount), and evolves only the dynamic characteristics. Although the method of causing the above has been described, this is not limited to the present embodiment, and a control module that combines static characteristics and dynamic characteristics in the evolution adaptive layer may be provided, and the entire combination may be evolved. Alternatively, control modules corresponding to static characteristics and dynamic characteristics may be independently provided in the evolution adaptation layer, and each may be evolved. Also,
When the control module corresponding to the static characteristic and the dynamic characteristic is independently provided in the evolution adaptive layer, these evolution processes may be performed in parallel, or one of them, preferably the static characteristic is evolved first, After that, the evolution process of the other control module to which the control module after evolution has been fixed may be performed.
Further, in the present embodiment, the evolution process of each control module is configured to end when the evolution of a predetermined number of generations is completed. However, the end determination of the evolution process is not limited to the present embodiment. May be configured to continuously perform the evolution processing of each control module until is converged. Further, in this embodiment, a push button type input device is provided, and the evaluation of each individual of the user is determined according to the length of time the user has pressed the button. The determination method is not limited to the present embodiment, and can be determined by any method.For example, during the evolution process, the characteristics of all individuals are visually displayed, and the user directly selects from them. You may comprise so that the individual with high evaluation may be selected.
Further, in the present embodiment, selection of individuals during the evolution process is performed according to the user's preference, but this is not limited to the present embodiment, and the use state of the vehicle, changes in the use environment, or Each individual may be evaluated based on the aging of the power source. Specifically, for example, the use state of a congested road, a mountain road, an expressway, city driving, or the like may be determined from the operation patterns of the throttle and the brake, and the evaluation criteria may be changed based on these. The evaluation criterion is, for example, appropriately set so that low response and high fuel efficiency have a higher evaluation on a congested road, and high response and low fuel efficiency have a higher evaluation on a mountain road. Can be changed. Also,
For example, the use environment may be determined based on the measured temperature, temperature, humidity, and atmospheric pressure, and the evaluation criteria may be changed based on the determination result. Further, for example, a change in the mounting state of the engine, wear of the cam of the camshaft, or a decrease in output due to wear of the piston ring, etc., are monitored to determine deterioration with time of the power source, and the evaluation criteria are determined based on the determination result. May be changed. In the embodiment described above,
Although the output control method for the power source mounted on the vehicle is described, the application object is not limited to this embodiment, and any vehicle with a power source may be used. Needless to say, it can be applied to

[0019]

As described above, the output control method for a vehicle with a power source according to the present invention directly outputs the power of the power source based on the operation amount of the primary output adjustment means operable by the user. In an output control method of a power source mounted on a vehicle with a power source for determining an operation amount of a secondary output adjustment unit to be adjusted, a relationship between an operation amount of the primary output adjustment unit and an operation amount of the secondary output adjustment unit is determined. Since it is evolved using a genetic algorithm based on at least one of the characteristics of the user, the state of use, a change in the use environment, or the aging of the power source, the amount of operation of the primary output adjustment means and the secondary output adjustment means At least the user's characteristics, usage conditions,
Since it can be adaptively changed according to either the change of the use environment or the aging of the power source,
For example, by evolving based on the characteristics of the difference in use, the output performance of a vehicle with a power source can reflect the taste and skill of the user, or the state of the user, and based on the use state and use environment. If it evolves, it will be possible to perform output control according to the use condition and use environment (such as weather and temperature of the area where it is used) of the powered vehicle,
If evolution is performed based on the aging of the power source, there is an effect that it is possible to perform optimal output control following the aging of the power source. In addition, by evolving the static characteristics between the operation amount of the primary output adjustment means and the operation amount of the secondary output adjustment means, it is possible to evolve the driving characteristics of the vehicle with the power source during steady running, and to improve the dynamic characteristics. By making the change, it is possible to improve the driving characteristics of the vehicle with the power source during transient operation. Furthermore, by evolving the primary delay and / or differential characteristics between the operation amount of the primary output adjustment means and the operation amount of the secondary output adjustment means, the response performance of the powered vehicle can be greatly improved. It has the effect of becoming. Further, by evolving the relationship between the amount of operation of the primary output adjustment means and the amount of operation of the secondary output adjustment means based on at least one of the user's preference, skill or state, the user's preference, skill or state Thus, it is possible to obtain an output characteristic suitable for the vehicle, and to provide a user with a pleasure of training a vehicle. Further, by providing a suitable input device that can be input by the user, based on the length of time the user operates this input device, by judging the user's preference,
Even if there is ambiguity in human evaluation, it is possible to obtain an evaluation value with a certain degree of accuracy, and it is possible to effectively use a genetic algorithm.

[Brief description of the drawings]

FIG. 1 is a schematic diagram showing a relationship between an engine and a control unit that executes an output control method.

FIG. 2 is a schematic block diagram showing an internal configuration of a control unit.

FIG. 3 is a graph showing an example of static characteristics of some throttles.

FIG. 4 is a graph showing examples of dynamic characteristics of some throttles.

FIG. 5 is a schematic block diagram of an evolution control unit.

FIG. 6 is a schematic block diagram illustrating a configuration of an electronic throttle control module.

FIG. 7 is a diagram illustrating a specific example of a control module in the evolution adaptive layer.

FIG. 8 is a flowchart showing a flow of an evolution process of a control module using a genetic algorithm.

FIG. 9 is a conceptual diagram showing an example of an individual obtained by coding a coupling coefficient of a neural network as a gene.

FIG. 10 is a diagram conceptually showing a method of acquiring teacher data used in a learning control module of a learning layer.

FIG. 11 is a diagram conceptually showing learning of a learning layer by CMAC using only new teacher data.

FIG. 12 is a flowchart schematically showing a flow of an entire evolution process in an evolution control unit.

FIG. 13 is a schematic diagram showing a relationship between an electric motor and a control unit that executes an output control method.

FIG. 14 is a schematic block diagram illustrating an internal configuration of a control unit.

FIG. 15 is a schematic block diagram illustrating an internal configuration of a motor power supply control unit.

FIG. 16 is a graph showing an example of static characteristics of some throttles.

FIG. 17 is a graph showing some dynamic characteristics when the electric motor is driven.

FIG. 18 is a graph showing some dynamic characteristics during electric motor regeneration.

FIG. 19 is a schematic block diagram of an evolution control unit.

──────────────────────────────────────────────────の Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) G05B 13/02 G05B 13/02 Z (72) Inventor Hiroaki Takechi 2500 Shinkai, Iwata-shi, Shizuoka Yamaha Motor Co., Ltd. In-house F-term (reference) 3G065 CA21 DA05 DA06 FA08 FA12 GA10 GA41 GA46 KA02 KA12 KA33 3G084 BA05 DA04 EB12 EC03 FA10 FA33 3G301 JA03 LA01 LC03 NA05 ND01 ND21 PA11A PE01Z PF03A 5H004 GA14 GA15 GA26 GB12 HA08 KB08 JA08 KB30 KB39 KD46 KD67 LA17 MA11 MA21

Claims (8)

[Claims] 1. A vehicle equipped with a power source for determining an operation amount of a secondary output adjusting means for directly adjusting an output of a power source based on an operation amount of a primary output adjusting means operable by a user. In the power source output control method, the relationship between the amount of operation of the primary output adjustment means and the amount of operation of the secondary output adjustment means, at least user characteristics, use state,
An output control method for a vehicle with a power source, wherein the evolution is performed using a genetic algorithm based on either a change in a use environment or a deterioration with time of the power source. 2. The relationship between the amount of operation of the primary output adjustment means and the amount of operation of the secondary output adjustment means is determined by the static characteristics and / or
The control method according to claim 1, wherein the control method is a dynamic characteristic. 3. The control method according to claim 2, wherein said dynamic characteristic includes at least one of a first-order lag and a differential characteristic. 4. The control method according to claim 1, wherein the characteristic of the user includes at least one of the preference, the skill, and the state of the user. 5. An apparatus according to claim 1, wherein an appropriate input device capable of inputting by the user is provided, and the user's preference is determined based on a length of time the user operates the input device. 5. The control method according to 4. 6. The control method according to claim 1, further comprising a device for operating the secondary output adjusting means based on an operation amount of the primary output adjusting means. 7. The power source is an engine, the primary output adjustment means is an accelerator pedal, an accelerator grip or an accelerator lever, the secondary output adjustment means is a throttle valve, and the secondary output adjustment means is 7. The control method according to claim 6, wherein the operating device is a stepping motor that drives the throttle valve. 8. The secondary power adjusting means, wherein the power source is an electric motor, the primary output adjusting means is one of an accelerator pedal, an accelerator grip, and an accelerator lever, the secondary output adjusting means is an electric motor, 7. The control method according to claim 6, wherein the device that operates is a control device that controls energization of the electric motor.