JP2013016075A - Information processor, information processing method and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To propose an information processor, etc. capable of obtaining a real solution by applying genetic algorithm (GA) even in the case of a large-scale and complicated road network.SOLUTION: In an information processor 1, gene information within an individual corresponds to a traffic signal in multiple traffic routes in a road network, and the traffic routes are arrayed in a multidimensional form within the individual. A genetic manipulation part 7 in the information processor 1 uses the multidimensional array to apply genetic manipulation to multiple genes. A correction part 9 in the information processor 1 prevents occurrence of fatal solution in each individual by correcting the gene information for each individual after the genetic manipulation in GA, in accordance with a rule stored in a constraint condition storage part 19.

Description

  The present invention relates to an information processing apparatus, an information processing method, and a program, and more particularly, to an information processing apparatus that performs information processing using a genetic algorithm.

  Chronic traffic congestion has a serious impact on the environment and economy. Traffic signals are expected to reduce traffic congestion by optimizing traffic signals. Traffic signal optimization is a difficult combinatorial problem. Therefore, application of a genetic algorithm (GA) which is one of combination optimization algorithms has been proposed (see Non-Patent Documents 1 and 2). Non-Patent Documents 1 and 2 describe that only a single road (see FIG. 8) is controlled using a GA. As shown in Fig. 8, a single road is connected by a single road from the entrance to the exit, and the roads intersecting at the intersections between them and their traffic conditions are defined locally. It is.

Sei Takahashi, 3 authors, “Searching for optimal traffic signal offset adapted to traffic flow fluctuations by genetic algorithm”, Electrical Engineering D, Vol. 123, No. 3, pp. 204-210, 2003. Tsutomu Shinkawa, 3 authors, “Comparison of meta strategies in signal control optimization and application to continuous automatic adjustment of control parameters”, Yamaguchi University Faculty of Engineering Research Report, vol.57, no.1, pp.21-24, 2006.

  However, in GA, genetic information within an individual is basically independent. Therefore, as described in Non-Patent Documents 1 and 2, it was basically applicable only to a single road with no traffic signal duplication. Therefore, control on a large and complex road network has not been achieved.

  Therefore, an object of the present invention is to propose an information processing apparatus and the like that can obtain a real solution by applying GA even to a large and complex road network.

  A first aspect of the present invention is an information processing apparatus for obtaining a plurality of control parameters of a traffic signal by a genetic algorithm, wherein each of a plurality of gene information in an individual includes a plurality of control parameters of a traffic signal, Restriction condition storage means for storing restriction conditions including a condition that part or all of the plurality of control parameters are the same for at least two gene information in the individual, and genes for the plurality of individuals Genetic manipulation means for performing manipulation, correction means for setting at least some of the plurality of control parameters to the same value for the at least two genetic information in each individual after genetic manipulation in accordance with the constraint, and after correction Evaluation means for evaluating each of the individual is provided.

  According to a second aspect of the present invention, in the first aspect, each of the genetic information corresponds to one of traffic signals in a plurality of one-way traffic routes, and the plurality of parameters are the traffic signals. The first type control parameter and the second type control parameter, wherein the constraint condition is the same value for the plurality of gene information corresponding to the same traffic signal in different traffic routes, The second type control parameter includes a condition that traffic signals corresponding to the same direction have the same value, and traffic signals corresponding to different directions have a sum less than or equal to a certain value. For a plurality of genetic information corresponding to the same traffic signal in different traffic routes, the first type control parameter is set to the same value, and the second type control parameter is set to the same value. It is intended to a sum less than a certain value.

  According to a third aspect of the present invention, in the second aspect, each individual is a multi-dimensional array corresponding to each traffic route, and the genetic manipulation means performs genetic manipulation on each dimension. .

  According to a fourth aspect of the present invention, in any one of the first to third aspects, each of the gene information corresponds to one of traffic signals in a plurality of one-way traffic routes, The parameter includes a plurality of control parameters of the traffic signal, and the constraint condition sets a part or all of the plurality of control parameters to the same value for genetic information corresponding to the traffic signal on the same traffic route. This condition is included.

  A fifth aspect of the present invention is an information processing method for obtaining a plurality of control parameters of a traffic signal by a genetic algorithm in an information processing device, wherein a plurality of pieces of gene information included in each of a plurality of individuals are respectively A plurality of control parameters for traffic signals, wherein the genetic manipulation means performs genetic manipulation on the plurality of individuals, and the correction means applies at least two pieces of genetic information to the plurality of control parameters. A correction step for correcting a part or all of the plurality of control parameters to the same value for the at least two gene information according to a constraint condition including a condition that part or all are the same, and an evaluation means, The evaluation step of evaluating each of the individual is included.

  A sixth aspect of the present invention is a program for realizing an information processing method according to the fifth aspect in a computer.

  The invention of the present application is an information processing apparatus that performs information processing using a genetic algorithm, and includes a constraint condition storage unit that stores constraint conditions for a plurality of pieces of genetic information in an individual, and performs genetic operations on the plurality of individuals. It may be understood as comprising a genetic manipulation means to be performed, a correction means for correcting a part of gene information in each individual after genetic manipulation in accordance with the constraint conditions, and an evaluation means for evaluating each individual after correction. .

  The present invention may be regarded as a computer-readable recording medium on which the program according to the sixth aspect is recorded (steadily).

  According to the present invention, it is possible for the inventors to realize optimization of traffic signals in a complex traffic network using a multi-element GA in which the GA is uniquely deepened. That is, in the conventional GA, genetic information in an individual is basically independent. Therefore, the conventional GA can only be applied to a single road with no traffic signal duplication. On the other hand, in the multi-element GA, by applying genetic information rules, it is possible to eliminate problems between genes, eliminate the occurrence of lethal solutions, and reliably derive real solutions. Therefore, it becomes possible to cope with a wide and complex road network.

  Furthermore, according to the present invention, it is possible to alleviate traffic congestion by optimizing various parameters of traffic signals. Examples of traffic signal control parameters include cycle time, split time, and offset. The cycle time is a time related to a change of “blue, yellow, red” and one round. The split time is a passage permission time assigned to each display of the signal. The offset is a deviation of the timing at which the display of the adjacent signal is switched. According to the present invention, it is possible to efficiently manage traffic by optimizing all control parameters. And, for example, if it is based on information distribution, etc., only those who receive the distribution will get profit. On the other hand, since the present invention uses a traffic signal that is an existing traffic infrastructure, it can produce profits for all drivers. Furthermore, it becomes possible to alleviate traffic congestion on a large scale.

  Furthermore, according to the fifth aspect of the present invention, by performing genetic manipulation with each dimension independently, each traffic route can be lethal within an individual while applying GA to a plurality of genes independently. By avoiding the solution, the parameter can be optimized more efficiently.

It is a schematic block diagram of the information processing apparatus which is an example of embodiment of this invention. It is a flowchart which shows an example of operation | movement of the information processing apparatus 1 of FIG. It is a figure which shows an example of the road network used as the object of a process. It is a figure which shows the example of the individual who performed the gene expression with respect to the road network of FIG. It is a figure which shows the traffic network used as the object of simulation. It is a figure which shows transition of the evaluation value by the number of generations about (a) simple roulette selection and (b) proposal roulette selection. It is a figure which shows the simulation result at the time of using a different objective function. In the optimization of a traffic signal, it is a figure which shows an example of the single road to which the conventional GA was applied.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. The present invention is not limited to this embodiment.

  FIG. 1 is a schematic block diagram of an information processing apparatus which is an example of an embodiment of the present invention. FIG. 2 is a flowchart showing an example of the operation of the information processing apparatus 1 of FIG. With reference to FIG.1 and FIG.2, an example of a structure and operation | movement of the information processing apparatus 1 is demonstrated.

  The information processing apparatus 1 performs information processing using a genetic algorithm (GA). GA is a probabilistic search method inspired by biological evolution. GA is an algorithm composed of three relatively simple genetic operations: selection, crossover, and mutation, and can be applied to all optimization and search problems. In GA, a group of virtual individuals with genes is set in the computer, and generational change simulations are performed so that individuals adapting to a predetermined environment have a high probability of leaving offspring. “Evolve” the group. In this way, the GA operates so as to generate an individual having a high fitness (a degree suitable for the environment).

  One of the major characteristics of general GA is the diversity of the applicable range. In other words, there are a wide range of problems that can be solved by applying GA, and furthermore, little knowledge of the problem is required. Because of these advantages, GA has been used for many optimization problems. However, on the other hand, due to the strong randomness, it is possible that an infeasible individual, that is, an individual that cannot be adapted to the environment given by the problem, may be generated in a complicated optimization problem with many constraints. . Such individuals are called lethal solutions. If this lethal solution occupies many individuals in the same generation, the number of individuals to be crossed decreases.

  The multi-element GA was originally developed by the inventors so as to solve the problem of lethal solutions in the GA, and to reliably change generations and derive real solutions even in complex optimization problems. It is. This multi-element GA has the following characteristics.

  First, it is multidimensional arrangement of individual expressions. In conventional GA, genetic manipulation is basically performed on only one element. However, multi-dimensional arrangement of individual representations has the advantage that GA can be adapted to more complex problems.

  Next, the rule is given to the gene information in the individual. In conventional GA, genetic information in an individual is basically independent. There are a lot of rules with the rule of “prohibiting the appearance of the same gene within an individual”, but it is clear that “if one gene is x, another gene must be y”. There is no GA with rules. By setting the rules for genetic information, there is an advantage that defects between genes can be eliminated, lethal solutions can be eliminated, and real solutions can be derived reliably.

  By using a multi-element GA having such characteristics, it is possible to apply a GA, which has conventionally been applicable only to a single road, to a wide and complex road network.

  Specifically, in this embodiment, traffic signal control parameters (cycle time, split time and offset) (an example of “control parameters” in the claims of the present application) are optimized in a wide and complex road network. In this embodiment, traffic management is performed more efficiently by optimizing all control parameters. In this way, when solving a complex optimization problem involving multiple elements by targeting multiple control parameters or expanding the controlled road network, when simply extending the conventional GA, The following situation occurs: That is, within an intersection where two roads intersect, a pair of traffic signals (traffic signals in one direction and another direction) show opposite indications. In addition, in the conventional GA, genetic manipulation of individuals such as crossovers and mutations cannot be performed between genes, and a lethal solution occurs.

  Therefore, the following rules are applied as these measures. In the road network to be optimized, the user himself / herself decides a plurality of a single road with a set of inlets and outlets, and subdivides the individual by that number (multidimensional expression) (rule 1). At the intersection where two or more roads intersect, the cycle time is made the same (rule 2-1), and the total split ratio (%) is made 100 (rule 2-2) (rules 2-1 and 2-2 are Also referred to as rule 2. Rule 2 is an example of “constraint condition” in the claims of the present application.) With these rules, a multi-element GA corresponding to traffic signal control is constructed. According to Rule 1, a real solution can be obtained more easily even in a complicated optimization problem. Furthermore, by applying rule 2 to each traffic signal, the individual is prevented from being lethal. These rules are stored in the constraint condition storage unit 19.

  The information processing apparatus 1 in FIG. 1 includes an initialization unit 3, an evaluation unit 5 (an example of “evaluation means” in the present application claim), a gene operation unit 7 (an example of “gene operation means” in the present application claim), , The correction unit 9 (an example of “correction means” in the claims of the present application), the control unit 11, the connection unit 13, the simulation unit 15, the individual set storage unit 17, and the constraint condition storage unit 19 (in the claims of the present application) An example of "constraint condition storage means". The gene manipulation unit 7 includes a selection unit 21 and a crossover / mutation unit 23.

  The individual set storage unit 17 stores a set of individuals to be processed by the GA. The constraint condition storage unit 19 stores constraint conditions regarding a plurality of pieces of gene information within an individual. A specific example of an individual and a constraint condition will be described with reference to FIGS.

  FIG. 3 shows a road network to be processed. FIG. 4 shows an example of an individual subjected to gene expression. In the road network of FIG. 3, rule 1 is applied and n user-defined single roads (the trunk line of FIG. 3) are determined. As shown in FIG. 4 (a), the individual is subdivided according to each single path. Here, like the signal Nos. 4 and 4 ', the genes expressed with' indicate the traffic signal pairs at the respective intersections.

  The initialization unit 3 in FIG. 1 applies rule 2 to each traffic signal in FIG. 4A to generate an initial individual set, and stores it in the individual set storage unit 17 (step ST1 in FIG. 2). . For example, as shown in FIG. 4B, for the signals 4 ', 5' and 12 ', the offsets and cycle times of the signals 4, 5 and 12 are matched. Then, the split ratio of the signals 4 ′, 5 ′, and 12 ′ is set to a value obtained by subtracting the split ratio of the signals 4, 5, and 12 from 100 (the control parameters of the offset and cycle time are “first parameter of claims”). It is an example of “seed control parameter”, and the split ratio is an example of “second type control parameter” in the claims of this application). This can prevent the individual from becoming lethal. When there is a road in the direction not to be processed at the intersection as shown by signal 5 in FIG. 3, the signal 5 and 5 ′ pair may have a split ratio of less than 100. Good.

  The evaluation unit 5 in FIG. 1 evaluates each individual stored in the individual set storage unit 17 (step ST2 in FIG. 2). The evaluation unit 5 and the simulation unit 15 are connected by a connecting unit 13 (API unit). The simulation unit 15 realizes a traffic flow by a traffic flow simulator. The evaluation unit 5 causes the simulation unit 15 to realize a traffic flow for each individual, and evaluates each individual using an objective function. One example will be specifically described.

The simulation unit 15 is realized by, for example, a high function traffic simulator Aimsun6. The simulation part 15 performs the simulation according to an actual traffic network by allocating the individual information produced | generated by GA with respect to the traffic network created in Aimsun6. The evaluation unit 5 takes in the simulation result and calculates the fitness. In order to calculate the fitness, the evaluation unit 5 first employs the vehicle passage ratio (Fitness), which is a simple objective function expressed by the equation (1), but may employ other objective functions. . Here, V _re (remaining vehicle amount) is the number of vehicles remaining in the road network. V _out (vehicle outflow amount) is the number of vehicles that have escaped from the road network. V _waitout (the number of vehicles waiting) is the number of vehicles waiting to _enter . These are the values at the end of the simulation. Equation (1) is the ratio of the number of vehicles that escape the road network. By using such a simple objective function, there is an advantage that the result can be easily understood visually. The higher this value, the better the solution. In this way, fitness is derived for each individual.

  The control unit 11 in FIG. 1 determines whether a specified number of generations has been processed (step ST3 in FIG. 2). GA is an algorithm that has no clear termination criteria. Therefore, it is necessary for the user to determine the end generation by checking the convergence state of the evaluation values. Here, from a preliminary experiment, for example, one generation is assumed to be 100 individuals, and 100 generations are regarded as end determination.

  The gene operation unit 7 in FIG. 1 performs gene operations on a plurality of individuals stored in the individual set storage unit 17. Specifically, the selection unit 21 in FIG. 1 selects some of the plurality of individuals stored in the individual set storage unit 17 based on the evaluation by the evaluation unit 5 (step ST4 in FIG. 2). The crossover / mutation unit 23 in FIG. 1 performs crossover by exchanging gene information in the selected individual and / or changes the gene information in the selected individual to mutate (FIG. 2). Step ST5).

The process by the selection part 21 is demonstrated concretely. Selection is a process of selecting individuals to be handed over to the next generation and individuals to perform genetic operations such as crossover and mutation. In the present embodiment, as the fitness, first, a simple objective function of Expression (1) is used. Simplification has the disadvantage that there is not much difference in fitness. Therefore, in this embodiment, the following selection is performed. First, individuals are ranked as 1, 2,..., K,. Next, power scaling is performed so that the rank k given to each individual is k ′ = k ² . The size of the roulette wheel is determined by the scaled k ′ ratio, and roulette selection is performed. In this way, ranking and scaling ensure that there is a clear difference in the individuals selected even in dense fitness. Furthermore, by utilizing the characteristics of roulette selection, all individuals have the potential to survive to the next generation.

  The processing by the crossover / mutation unit 23 will be specifically described. A genetic operation such as crossover and / or mutation is performed on the individual selected by the selection unit 21 to create a next generation individual. In this embodiment, as the crossover method, the one-point crossover method which is the simplest method is used. The mutation method uses a method of randomly mutating a gene selected according to the mutation rate. In addition, general values are used for the crossover rate and mutation rate.

  The correction unit 9 in FIG. 1 corrects part of the gene information in the individual by applying the rule 2 stored in the constraint condition storage unit 19 to each individual after the genetic manipulation (in FIG. 2). Step ST6). Specifically, the operation of the initialization unit 3 is the same as that described with reference to FIG. And it returns to the process of step ST2 of FIG.

  FIG. 4 is an example in which traffic signal pairs have different directions at intersections. However, traffic signals in the same direction may appear on different single roads. In this case, it is possible to set the rule 2 that all the control parameters have the same value, and to correct them by the correcting unit 9. Further, as rule 2, for example, the cycle time of traffic signals on the same traffic route may be made constant.

  Subsequently, in the present embodiment, it is confirmed by comparing the theoretical optimum solution and the simulation result whether the multi-element GA converges with the actual solution and operates as expected. Here, as shown in FIG. 5, a simple traffic network that can predict the result is created. That is, as the road structure, the roads A and B and the roads α and β are orthogonal to each other, and the road length is the same from the vehicle inlet to the intersection inlet, between the intersections, and from the intersection outlet to the vehicle outlet. (200m). In each road, the opposite direction is indicated by attaching a symbol “′”. The number of vehicles is 900 cars / h for roads α and A, 300 cars / h for road β, 100 cars / h for road B, and 0 cars / h for roads α ′, β ′, A ′ and B ′. . Table 1 shows the simulation conditions.

  FIG. 6 is a diagram showing transition of evaluation values according to the number of generations for (a) simple roulette selection and (b) proposed roulette selection. In (a), although the average value and the minimum value are increasing, the maximum value is decreasing. On the other hand, in (b), although the minimum value fluctuates, the average value and the maximum value almost increase. Therefore, it can be said that (b) the proposed roulette selection ensures that the real solution of each generation is inherited to the next generation and promotes evolution between generations. This is because even if similar fitness levels are concentrated by ranking and scaling each individual, the difference in fitness level is clarified, so that the individual with higher fitness level can be selected. This is thought to be due to the rise. In addition, the fitness has converged toward the final generation from the distribution of each generation, and it can be said that this embodiment operates as expected.

  Table 2 shows the amount of vehicle inflow from each inflow port, vehicle acceleration, speed limit, speed limit value (speed limit value exceeds the speed limit on roads where speed limit is set in the real world) This is a parameter that generates a vehicle that travels at a speed that is a multiple of the speed limit from the speed limit, taking into account the vehicle.) Calculated from deceleration, a formula provided by the Traffic Engineering Society, etc. The theoretical value is shown. Table 3 shows actual solutions of control parameters obtained by simulation. Comparing Table 2 and Table 3, the theoretical optimal solution and the actual solution of the present example are close to some extent, and the effectiveness of the present example is shown.

  In order to realize traffic signal control by a complex and wide road network, it is necessary to select an appropriate objective function. The objective function of Equation (1) converges without initial convergence (see FIG. 6), and it can be seen that this example shows correct behavior. However, the objective function of Equation (1) considers only the number of vehicles at the end of the simulation. Therefore, there is a drawback that a relatively good result is output regardless of the traffic parameter.

Therefore, in the following, to solve this problem, an example using an objective function that takes into account the physical quantity in units of speed and time that can be considered even during simulation, with a wide range of fitness is explained. To do. Specifically, it is represented by the formula (2). Here, V _S is a simulation evaluation, and V _E is an emotion evaluation. It is assumed that the fitness derived from the objective function V is different from usual, and a lower value is an excellent evaluation with a higher degree of adaptation to the environment.

The objective function V _S in the simulation evaluation adopts Delay Time. The delay time is the time difference between the travel time required when the vehicle travels under ideal conditions and passes through the traffic network and the actual travel time. The reason is that it is possible to consider the delay time caused by stepping on the brake without stopping before the intersection due to offset, the data obtained during the simulation unlike the vehicle passing rate, and the change in each control parameter in the simulation results. It makes a big difference. The range of values for the delay time varies greatly depending on the vehicle length and the inflow volume. Therefore, the range can be adjusted by a constant. From the simulation results, the ratio of the maximum value to the minimum value in the offset was about 1.3 times. For this reason, an exponential function is used as shown in Equation (3) in order to provide more flexibility. Here, t _Delay is Delay Time (s), and C is a constant. C is set such that the ratio of Delay Time is at least doubled by setting C = 50.

In addition, if the cycle takes a large value, the intersection demand rate also becomes high, the Delay Time is low, and it is stabilized at a constant value. Therefore, if a real solution is found in consideration of only traffic conditions, the cycle should be increased. For this reason, when the Delay Time becomes a low constant value, there are many similar individuals, and a single solution cannot be given. Therefore, it is necessary to give a bad evaluation so that the cycle does not grow indefinitely. In this case, the waiting time due to the red display also becomes long, and a mental burden is generated on the driver waiting for the blue display. So, as an objective function V _E in the emotion evaluation, such as cycle it is and give a bad evaluation long, has introduced the section, including the emotions. As the driver's feelings, the increase in the stop time when the stop time due to the red display is short does not cause much mental burden. However, when the stop time is long (the cycle is large), even if the stop time is the same as when the stop time is short, the driver's mental burden increases. Therefore, emotion evaluation V _E, as in equation (4), using an exponential function such that a large value enough to cycle becomes longer. Here, C _a and C _b are constants and adjust emotion evaluation. The cycle changes with 40 seconds as the minimum, and Ca = 40 is set so that the minimum is 1 in the emotion evaluation. Also, since there are many traffic signals that use 70 seconds or 140 seconds as the cycle, Cb = 100 is set so that a difference of about 2 times appears.

  FIG. 7 shows a simulation result using Equation (2) as the objective function. FIG. 7 shows the minimum, maximum and average fitness in each generation. From FIG. 7, it can confirm that it has converged correctly and shows correct behavior. In addition, the optimal cycle could be derived uniquely when using equation (2) as the objective function. Furthermore, in the control parameter indicating the theoretical optimal solution, the fitness was 4.572223 when the fitness was calculated by the objective function of Equation (2). On the other hand, the fitness in the real solution obtained by the simulation using the equation (2) was 4.572225. From this, it can be said that the proposed objective function shown in Equation (2) is useful.

  DESCRIPTION OF SYMBOLS 1 Information processing apparatus, 3 Initialization part, 5 Evaluation part, 7 Gene operation part, 9 Correction part, 11 Control part, 13 Connection part, 15 Simulation part, 17 Individual set storage part, 19 Restriction condition storage part, 21 Selection part , 23 Crossover / mutation section

Claims (6)

An information processing apparatus for obtaining a plurality of control parameters of a traffic signal by a genetic algorithm,
Each of the plurality of genetic information in the individual includes a plurality of control parameters of traffic signals,
Restriction condition storage means for storing a restriction condition including a condition that part or all of the plurality of control parameters are the same for at least two gene information in the individual,
Genetic manipulation means for performing genetic manipulation on a plurality of the individuals,
In accordance with the constraint condition, with respect to the at least two gene information in each individual after genetic manipulation, a correction means for making some or all of the plurality of control parameters the same value,
An information processing apparatus comprising evaluation means for evaluating each individual after correction. Each of the genetic information corresponds to one of traffic signals in a plurality of one-way traffic routes,
The plurality of parameters include a first type control parameter and a second type control parameter of the traffic signal,
The constraint condition is for a plurality of genetic information corresponding to the same traffic signal in different traffic routes,
The first type control parameter has the same value,
The second type control parameter includes a condition that traffic signals corresponding to the same direction have the same value, and traffic signals corresponding to different directions have a sum less than a certain value,
The correction means sets the first type control parameter to the same value and sets the second type control parameter to the same value for a plurality of gene information corresponding to the same traffic signal in different traffic routes according to the constraint condition. The information processing apparatus according to claim 1, wherein the total sum is equal to or less than a certain value. Each individual is a multidimensional array according to each traffic route,
The information processing apparatus according to claim 2, wherein the gene operation means performs a gene operation for each dimension. Each of the genetic information corresponds to one of traffic signals in a plurality of one-way traffic routes,
The plurality of parameters includes a plurality of control parameters of the traffic signal,
The constraint condition includes a condition that a part or all of the plurality of control parameters have the same value with respect to genetic information corresponding to traffic signals on the same traffic route. The information processing apparatus described. In an information processing apparatus, an information processing method for obtaining a plurality of control parameters of a traffic signal by a genetic algorithm,
Each of the plurality of gene information included in each of the plurality of individuals includes a plurality of control parameters of traffic signals,
A genetic manipulation step, wherein the genetic manipulation means performs genetic manipulation on the plurality of individuals;
According to a restriction condition including a condition that a part or all of the plurality of control parameters is the same for at least two pieces of gene information, the correcting unit applies one of the plurality of control parameters to the at least two pieces of gene information. A correction step for correcting a part or all to the same value;
An information processing method including an evaluation step in which an evaluation means evaluates each individual after correction.   The program for implement | achieving the information processing method of Claim 5 in a computer.