CN112116207A - Multi-constraint-condition train operation adjustment calculation method and device - Google Patents

Abstract

The embodiment of the invention provides a method and a device for calculating train operation adjustment under multiple constraint conditions, wherein the method comprises the following steps: s1, modeling the high-speed train operation adjustment problem based on the genetic algorithm; s2, constructing an optimization target comprehensively considering the late time and the energy consumption of the train according to the established model; and S3, under the constraint condition, adjusting the train according to the optimization target. The invention can shorten the time of the later point as much as possible under the condition of ensuring the safe tracking distance of the train and the power supply.

Description

Multi-constraint-condition train operation adjustment calculation method and device

Technical Field

The invention relates to the field of traffic, in particular to a calculating device for train operation adjustment under multiple constraint conditions.

Background

The high-speed train is used as an important component of rail transit and transportation and bears important development strategies in the aspects of national economic development, cultural exchange, urban construction and the like. As a main mode of China railway passenger transportation, passengers with more than 90 hundred million people are transported to the south and north of the great river by high-speed trains, and the rapidness and the safety become the mark of the method. However, when adverse forces such as severe weather, equipment failure, artificial emergencies and the like occur in the operation environment, the train running at high speed is forced to slow down or even stop and a late point occurs, and because the train runs at high speed, the number of trains is large, the running interval is small, and the late point is quickly spread on a line, a large number of trains are delayed, and great threat is generated to the quick and stable operation of the high-speed train. After the interference is eliminated, each delayed train cannot run according to the original plan, and how to rearrange the running plan of the train can restore the original order of the line as soon as possible, shorten the late time and have important significance for ensuring the safe and efficient running of the high-speed train.

At present, many studies are made at home and abroad on adjustment of a high-speed train operation plan. Early part of researchers establish a mathematical model for solving through operation research methods such as a mathematical programming method, a branch definition method and the like, but the model is complex and the search space is large, so that the optimal solution is difficult to obtain; with the rapid development of computer technology and artificial intelligence methods, intelligent algorithms are gradually applied to the research of train operation adjustment problems, common algorithms comprise genetic algorithms, ant colony algorithms, particle swarm algorithms, firefly algorithms, simulated annealing algorithms and the like, the solution optimization targets are the total delay time of all trains, the delay time of the terminal station, the weighted delay time after different weights are divided according to different train speed grades and the like, the solution of the problems is expressed in the form of the departure time and the arrival time of each train at each station, and the solution meeting the conditions can be obtained as an adjustment plan after the delay time by establishing a model and using a computer to solve.

However, the existing method has some defects: firstly, the scheduling problem of the high-speed train is not an ideal independent existence, when dispatching and arrival time are arranged, the speed change of the train in the interval running needs to be considered, the speed limit of the special sections such as the gradient, the curve and the bridge tunnel of the line per se and the speed limit of the following train after the dispatching interval is shortened, the safe tracking distance is guaranteed not to exceed due to the influence of the position of the preceding train, namely, the speed limit is an important constraint condition in the algorithm solving process; secondly, the high-speed train provides power by means of electric power, when the number of trains in one interval is large, the power supply is insufficient, so that the trains cannot accelerate, namely the power supply capacity is another constraint condition; in addition, from the energy-saving perspective, if the arrangement is reasonable, the energy consumption in the operation process can be reduced, and the energy consumption can be used as another index for measuring the adjustment scheme except the late time.

With the rapid development of rail transit and the increase of travel demands of people, the number of trains is continuously increased, the running speed is continuously improved, the running interval of adjacent trains is smaller and smaller, and higher requirements are provided for controlling the speed of the trains and ensuring the safety. The high-speed railway adopts an ATP speed protection technology to monitor the running speed of the train, an ATP computer obtains the position information of the train in front and the line information of a slope, a curve, a bridge tunnel, a special section and the like, and the maximum speed limit allowed by the train at the current position is calculated by combining the basic parameters of the train. Taking two running trains as an example, the train control system sends the position of the front train to the rear train, and for the rear train, the position of the front train is a dangerous point, so that the speed limit of the rear train is calculated, the position of the front train is continuously changed, the speed limit of the rear train is changed along with the position of the front train, for the front train, the front station is a stopping point, the distance between the front train and the stopping point can be used for calculating the speed limit of the front train, a train speed-distance mode curve can be obtained, the actual running speed of the train must strictly follow an ATP curve, and the speed limit value cannot be exceeded.

The high-speed train runs by electric drive, and a traction power supply system of the high-speed railway needs to ensure safe, reliable and uninterrupted stable power transmission. Because the electric power that the contact net provided is usually the fixed value, when the train operation interval is shorter, electric power system probably can not guarantee that all trains obtain sufficient electric energy.

Disclosure of Invention

The embodiment of the invention provides a method and a device for calculating train operation adjustment under multiple constraint conditions, which can shorten the delay time as far as possible under the condition of ensuring the train safety tracking distance and the electric energy supply.

A method for calculating multi-constraint train operation adjustment comprises the following steps:

s1, modeling the high-speed train operation adjustment problem based on the genetic algorithm;

s2, constructing an optimization target comprehensively considering the late time and the energy consumption of the train according to the established model;

and S3, under the constraint condition, adjusting the train according to the optimization target.

A multi-constraint train operation adjustment computing device, comprising:

the modeling unit is used for modeling the high-speed train operation adjustment problem based on the genetic algorithm;

the building unit is used for building an optimization target comprehensively considering the late time and the energy consumption of the train according to the built model;

and the adjusting unit adjusts the train according to the optimization target under the constraint condition.

The invention researches the train operation adjustment problem based on multiple constraint conditions and multiple optimization targets, effectively combines the adjustment scheme with the train operation control, realizes the integration of dispatching and train control, and can shorten the late time as much as possible under the condition of ensuring the train safe tracking distance and the electric energy supply.

As can be seen from the technical solutions provided by the above embodiments of the present invention, in the embodiments of the present invention,

additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without inventive labor.

FIG. 1 is a schematic diagram of a method of calculating a multi-constraint train operation adjustment of the present invention;

FIG. 2 is a schematic diagram of a train target-distance model of the present invention;

FIG. 3 is a schematic representation of a train traction characteristic of the present invention;

fig. 4 is a schematic diagram of train traction power supply of the invention.

FIG. 5 is a schematic diagram of a first generation population of the present invention.

FIG. 6 is a diagram of the seventh generation population of the present invention.

FIG. 7 is a schematic representation of the total late time course of the genetic process of the present invention.

FIG. 8 is a diagram illustrating total energy consumption variation in the genetic process of the present invention.

FIG. 9 is a diagram illustrating the variation of fitness value in genetic processes according to the present invention.

Detailed Description

Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the accompanying drawings are illustrative only for the purpose of explaining the present invention, and are not to be construed as limiting the present invention.

For the convenience of understanding the embodiments of the present invention, the following description will be further explained by taking several specific embodiments as examples in conjunction with the drawings, and the embodiments are not to be construed as limiting the embodiments of the present invention.

As shown in fig. 1, the method for calculating the adjustment of train operation under multiple constraint conditions according to the present invention includes:

s1, modeling the high-speed train operation adjustment problem based on the genetic algorithm;

s2, constructing an optimization target comprehensively considering the late time and the energy consumption of the train according to the established model;

and S3, under the constraint condition, adjusting the train according to the optimization target.

The step 3 comprises the following steps:

s31, representing the train operation adjustment plan in the form of train operation interval time;

s32, the train runs at full speed according to the maximum speed limit of the line for the first time, the maximum speed limit is sequentially calculated according to the position of the train and the position of the preceding train determined by the interval time of each subsequent time, a speed-position curve is generated, and the interval running time is determined;

s33, determining the actual arrival time according to the departure interval time and the interval running time of each train;

s34, determining the late time by combining the plan operation chart;

s35, calculating energy consumption according to the speed-position curve;

s36, weighting and summing the late time and energy consumption to obtain a target function as a basis for determining the quality degree of the adjustment scheme;

s37, entering the optimizing process of the population according to the selection, crossing and variation of the genetic algorithm;

and S38, judging whether the algorithm reaches the convergence termination condition, if not, jumping to the step S31, otherwise, ending the processing, and taking the current result as the optimal adjustment scheme.

The step 3 comprises the following steps:

step 31, generating a first generation population by a random number generating mode through a computer, wherein population individuals are larger than a preset threshold value so as to ensure randomness and facilitate subsequent calculation;

step 32, among the trains participating in the adjustment, if no other train in front of the train which starts for the first time is a tracking point, taking a station which arrives in front as the tracking point, accelerating to the maximum speed limit of the line according to the maximum traction force, keeping cruising, and calculating a speed-position curve in a mode that the maximum braking force is decelerated to the station in front to stop, taking a corresponding time step to improve the calculation accuracy, and storing the speed, the position, the time, the traction current and the voltage value of the train at each moment;

step 33, calculating the maximum allowable speed limit at each moment by using the train closest to the front as a tracking point for each train in subsequent running, obtaining a speed-position curve according to the speed limit, and when the departure interval is small, the speed limit of the train in the rear running is easily influenced by the front train and reduced; the method comprises the steps that each train is dispatched according to dispatching interval time, the position of a front vehicle is obtained at each moment, the maximum speed limit is calculated, each train is guaranteed not to exceed the speed limit to run, and the speed, the position, the time and the traction current value of each train are recorded at each moment;

step 34, obtaining the total running time of each train according to the speed-position curve, and determining the transition point of the traction and cruising working conditions and the running distance under each working condition;

step 35, calculating the arrival time according to the interval running time of each train and the departure time:

step 36, the difference between the actual arrival time and the planned arrival time is the time of the train at the later point, the sum of the time of the train at the later point is the total time of the later point, the energy consumption is the integral of the traction force to the displacement under the traction and cruise working conditions, and the time of the later point and the energy consumption are normalized according to different weights and then summed to obtain the fitness value:

Ψ₁＝q_T*T_Delay+q_E*E (4)

step 37, obtaining the partial pressure of the contact network in the current power supply interval according to the traction current and displacement data in the running process of each train, judging whether the power supply constraint condition is met, and if the power supply constraint condition is not met, adding one to the fitness value to serve as the final fitness;

Ψ₂＝Ψ₁+sgn(U) (5)

and step 38, entering next generation operation and repeating the processes from step 32 to step 37.

The calculation mode of the fitness comprises the following steps:

(1) decoding the individuals to obtain the departure interval time of each train participating in adjustment;

(2) determining departure time of each train according to departure interval time by referring to a timetable;

(3) according to the ATP speed limit constraint condition, sequentially calculating the speed, the position, the traction current, the partial pressure and other data of each subsequent train in the running process according to the position of the front train;

(4) judging whether the electric energy meets the power supply capacity constraint or not according to the required electric energy of the high-speed train by the actual current and voltage value;

(5) determining the actual arrival time of each train according to the running time of each train;

(6) calculating the late time according to the formula (8) according to the planned arrival time and the actual arrival time,

is the actual arrival time of the ith train,

the planned arrival time of the ith train;

(7) calculating the total energy consumption of each train in a mode of integrating the traction force with the displacement according to the working condition in the running process according to the formula (9), wherein F_i,TraThe magnitude of the traction force of the ith train under the traction working condition, s_i,TraDistance traveled under traction condition of i-th train, F_i,CruThe magnitude of the traction force s under the cruising condition of the ith train_i,CruThe cruising working condition is the cruising working condition lower running distance of the ith train;

(8) normalizing the total late time and the total energy consumption to obtain a fitness function, calculating according to a formula (10), and q_TWeight in the objective function for the late timeHeavy, q_EFor the weight occupied by energy consumption in the objective function, T_MinTo minimum time of night, T_MaxTo maximum evening time, E_MinTo minimize energy consumption, E_MaxPsi is the fitness value for maximum power consumption, sgn (U) is a function that determines whether the power supply capability constraint is met, U_min、U_maxRespectively obtaining a minimum value and a maximum value of voltage required by normal operation, if the minimum value and the maximum value are not satisfied, obtaining a value of M, if the minimum value and the maximum value are satisfied, obtaining a value of 0, wherein M is a proper positive number, and selecting M as 1 as the fitness function value is between 0 and 1;

the specific steps of the ATP speed limit constraint include:

on the basis of a train dynamic model, a maximum allowable speed of a rear train is obtained according to the departure interval time of the train by combining a traction characteristic curve and other related parameters of the train, so that a speed-position curve of each train is determined, and the actual running time of the train is calculated;

the specific steps of the power supply capacity constraint judgment comprise:

and determining the traction current and the partial pressure of each train according to the speed-position relation in the running process of the trains, and further judging whether the adjustment scheme meets the constraint of the power supply capacity.

The parameters of the genetic algorithm are determined according to the following steps that three main genetic operations of selection, crossing and mutation are contained in the genetic algorithm, and through the operation, an individual can complete the optimal solution of the problems of high-quality elimination and gradual approximation; the selection means that the solution with higher quality directly enters the next generation by copying, the crossover finger exchanges partial genes of two chromosomes, the variation means that partial genes of one chromosome are randomly changed, and the crossover and variation are both ways of generating new individuals and searching the solution with higher quality. And each genetic operation needs to be completed according to a certain probability, and the specific calculation mode is as follows:

in formulae (20) and (21), P_cRepresenting the cross probability, P_mRepresenting the probability of variation, k_cRepresenting the cross probability base value, k_mRepresenting the base value of the probability of variation, f_maxRepresents the maximum value of individual fitness in the population, f_cRepresenting the greater fitness of the two chromosomes undergoing crossover operation, f_mRepresenting the fitness value of the chromosome on which the mutation was performed,

represents the average of fitness of all chromosomes in the population.

The probability of crossing and mutation of the selected individuals with high adaptability is higher, otherwise, the probability of crossing and mutation of the selected elite individuals with low adaptability is lower, and the elite individuals with low adaptability generally enter the next generation population directly through selection and are set with corresponding probabilities, so that the population is promoted to develop gradually towards the direction of higher solution quality by the three genetic operations.

The termination conditions include: reaching the maximum iteration times; the longest calculation time is exceeded; and thirdly, converging the population.

The judgment condition of the condition (c) is as follows:

in the formula (f)_iIs the fitness of the ith chromosome and is,

is population average fitness and is a small constant. If the population fitness satisfies the above formula, it can be considered that the population has reached convergence, the individuals in the population are already the optimal solution, the original individuals are not changed any more by continuing the operation, and at this time, the operation can be ended and the optimal solution obtained by the calculation can be output.

The invention also provides a calculating device for train operation adjustment under multiple constraint conditions, which comprises:

the modeling unit is used for modeling the high-speed train operation adjustment problem based on the genetic algorithm;

the building unit is used for building an optimization target comprehensively considering the late time and the energy consumption of the train according to the built model;

and the adjusting unit adjusts the train according to the optimization target under the constraint condition.

The following describes an application scenario of the present invention. In order to achieve the purpose of restoring the line operation order as soon as possible, the invention adopts a genetic algorithm model with multiple constraint conditions and multiple optimization targets, considers the influence of train speed limit and power supply in a multi-train tracking operation scene, formulates an optimization target with the minimum total delay time and energy consumption, and designs a train operation optimal adjustment strategy calculation method based on the genetic algorithm. The adjusting scheme provided by the invention can shorten the time at the night as far as possible under the condition of ensuring the safe tracking distance of the train and the power supply, and has important practical significance.

A train operation adjustment scheme calculation method considering multiple constraint conditions comprises the following steps:

s1: modeling a high-speed train operation adjustment problem based on a genetic algorithm;

s2: constructing an optimization target comprehensively considering the late time and the energy consumption of the train;

s3: the calculation solving method of the design consideration algorithm under the constraint condition comprises the following specific steps:

s31: representing the train operation adjustment plan in the form of train operation interval time;

s32: the train runs at full speed according to the maximum speed limit of the line for the first time, the maximum speed limit is sequentially calculated according to the position of the train determined and the front train at intervals for the subsequent times, a speed-position curve is generated, and the interval running time is determined;

s33: determining the actual arrival time according to the departure interval time and the interval running time of each train;

s34: determining the late time by combining the plan operation diagram;

s35: calculating energy consumption from the velocity-position curve;

s36: weighting and summing the late time and the energy consumption to obtain a target function as a basis for determining the quality degree of the adjustment scheme;

s37: entering a population optimizing process according to selection, crossing and variation of a genetic algorithm;

s4: and repeating the steps S31-S36 until the algorithm reaches convergence to find out an optimal adjustment scheme.

Secondly, the modeling process of the high-speed train operation adjustment problem is as follows:

1. model parameters and variables

In order to describe and calculate the problem conveniently, parameters and variables used for solving and calculating are listed in this subsection:

m: adjusting the total number of stations contained in the interval;

n: the total number of trains running in the interval;

the planned arrival time of the ith train at the jth station;

the planned departure time of the ith train at the jth station;

actual arrival time of the ith train at the jth station;

actual departure time of the ith train at the jth station;

t_i,j: the operation time of the ith train in the jth interval;

T_i,i+1: the departure interval time of the ith train and the (i + i) th train;

v_max: the maximum speed limit of the line;

t_i,j,Tra: the running time of the ith train under the traction working condition in the jth interval;

t_i,j,Cru: the running time of the ith train under the cruising working condition in the jth interval;

s_i,j,Tra: the running distance of the ith train under the traction working condition in the jth interval;

s_i,j,Cru: the running distance of the ith train under the cruising working condition in the jth interval;

E_i,j,: the ith train runs energy consumption in the jth interval;

U_j: the voltage of a contact network in the jth power supply interval;

2. assumption of model conditions

According to the actual characteristics of the high-speed train operation and the relevant actual conditions of the railway line, the following assumed conditions are made for the convenience of analysis and calculation:

(1) the emergency only affects a single direction, and the solution of the optimal adjustment plan also aims at a single-direction train;

(2) the actual speed limit value of each train is the smaller value of the maximum speed limit of the line and the speed limit calculated by the position of the train running in front;

(3) the running process of the high-speed train is based on the principle of traction-cruise-braking, the idling working condition is not considered generally, and the condition of multiple conversion of the traction-cruise working condition is also not considered;

(4) the calculation of traction current, partial pressure and energy consumption in the power supply system only aims at traction and cruising working conditions, and the influence of regenerative braking current is not considered;

(5) the departure interval value of each train is an integer and the unit is minute.

3. Genetic algorithm model

From the above analysis, a specific solution model for the genetic algorithm can be determined as follows:

(1) generating a first generation population by a computer to generate random numbers, wherein population individuals need to be sufficient to ensure randomness and facilitate subsequent calculation;

(2) among the trains participating in adjustment, if no other train in front of the train which starts for the first time is taken as a tracking point, the arriving station in front is taken as the tracking point, the cruise can be kept after the maximum traction force is accelerated to the maximum speed limit of the line, the speed-position curve is calculated in a mode that the maximum braking force is decelerated to the stop of the station in front, the proper time step length is taken to improve the calculation accuracy, and the speed, the position, the time, the traction current and the traction voltage value of the train at each moment are stored;

(3) the train in each time of subsequent running uses the nearest train in front as a tracking point, calculates the maximum allowable speed limit at each moment, obtains a speed-position curve according to the speed limit, and when the departure interval is smaller, the speed limit of the train in the back is easily influenced by the front train and is reduced. Then, sequentially arranging each train for departure according to departure interval time, acquiring the front position at each moment to calculate the maximum speed limit, ensuring that each train does not exceed the speed limit, and recording the speed, position, time and traction current value of the train at each moment;

(4) the total running time of each train can be obtained according to the speed-position curve, and the transition point of the traction working condition and the cruising working condition and the running distance under each working condition are determined;

(5) the arrival time can be calculated by combining the interval running time of each train with the departure time:

(6) the difference value of the actual arrival time and the planned arrival time is the late time of the train, the sum of the late time of each train is the total late time, the energy consumption is the integral of the traction force to the displacement under the traction and cruise working conditions, and the late time and the energy consumption are normalized according to different weights and then summed to obtain the fitness value:

Ψ₁＝q_T*T_Delay+q_E*E (4)

(7) the magnitude of the partial pressure of the contact network in the current power supply interval can be obtained from the traction current, the displacement and other data in the running process of each train, whether the power supply constraint condition is met or not is judged, and if the condition is not met, the fitness value is increased by one to serve as the final fitness;

Ψ₂＝Ψ₁+sgn(U) (5)

(8) and (5) entering next generation operation and repeating the processes (2) to (7).

4. Optimization target of high-speed train operation adjustment problem

The research objective of the high-speed train operation adjustment problem is to determine an optimal scheme which enables the optimal objective to reach the maximum value, namely, an objective function is determined to evaluate the quality of the problem solution. In the genetic algorithm, the fitness function can be used for describing the quality degree of the obtained solution, and the fitness is used as a judgment standard to determine whether the individual is directly selected to enter the next generation of population or generate a new individual through crossing and variation. Among the research problems of this patent, the optimization target is: the time of all the late points of each train participating in adjustment is shortest, the energy consumed in operation is smallest, the quality of the solution is negatively related to the fitness, namely the smaller the fitness value is, the higher the individual quality is, the closer to the optimal solution of the problem is, and the objective function is normally positively related to the quality of the solution, so that the objective function value can be obtained by using a proper positive number to make a difference with the fitness value. In addition, the unit of the late time is minutes, the unit of the energy consumption is kJ, the magnitude and the unit of the two indexes are different, and the two indexes cannot be directly added, so that the two indexes need to be firstly transformed by a normalization method, the result is described as a value between 0 and 1, and then the fitness is converted into a target function, and the method comprises the following specific steps:

(1) decoding the individuals to obtain the departure interval time of each train participating in adjustment;

(2) determining departure time of each train according to departure interval time by referring to a timetable;

(3) according to the ATP speed limit constraint condition, sequentially calculating the speed, the position, the traction current, the partial pressure and other data of each subsequent train in the running process according to the position of the front train;

(4) judging whether the electric energy meets the power supply capacity constraint or not according to the required electric energy of the high-speed train by the actual current and voltage value;

(5) determining the actual arrival time of each train according to the running time of each train;

(6) calculating the late time according to the formula (8) according to the planned arrival time and the actual arrival time,

is the actual arrival time of the ith train,

the planned arrival time of the ith train;

(7) the total energy consumption of each train is calculated by integrating the traction force with the displacement according to the working conditions in the running process according to the formula (9)In (F)_i,TraThe magnitude of the traction force of the ith train under the traction working condition, s_i,TraDistance traveled under traction condition of i-th train, F_i,CruThe magnitude of the traction force s under the cruising condition of the ith train_i,CruThe cruising working condition is the cruising working condition lower running distance of the ith train;

(8) normalizing the total late time and the total energy consumption to obtain a fitness function, calculating according to a formula (10), and q_TWeight in the objective function for the late time, q_EFor the weight occupied by energy consumption in the objective function, T_MinTo minimum time of night, T_MaxTo maximum evening time, E_MinTo minimize energy consumption, E_MaxPsi is the fitness value for maximum power consumption, sgn (U) is a function that determines whether the power supply capability constraint is met, U_min、U_maxThe voltage is respectively the minimum value and the maximum value of the voltage required by normal operation, if the voltage is not satisfied, the value is M, if the voltage is satisfied, the value is 0, M is a proper positive number, and since the fitness function value is between 0 and 1, M can be selected as 1.

5. Constraint condition of high-speed train operation adjustment problem

(1) ATP rate-limiting constraint

In the research of train operation adjustment problem, the shortest total delay time of each train is an important optimization target, and in order to achieve the purpose of eliminating delay as soon as possible, the train which has generated delay needs to be dispatched as soon as possible, therefore, shortening the dispatching interval time of the delay train is the most direct means, but in the actual arrangement of the adjustment scheme, usually only the arrival and dispatching time of the delay train at each station is determined, and the operation details of the train in each section are not concerned, in the train operation control system, in order to ensure the safe operation of the train, the ATP speed protection technology must be adopted to accurately monitor the train operation speed in real time, when the distance between two adjacent trains is smaller, in order to ensure the safety, the actual allowable speed of the train in the rear row and the train is lower than the maximum speed limit of the line, the train can not operate according to the maximum speed limit, and the operation time is increased, which contradicts the target of shortening the delay time, therefore, how to make the train arrive at the station as early as possible is very important on the premise of ensuring the safety, so in the algorithm solving process of the patent, the ATP speed limit is one of the important constraint conditions. The ATP computer obtains the dynamic information and the basic route information of the front train and calculates the maximum speed limit allowed by the train at the current position by combining the basic parameters of the train. Taking two running trains as an example, the train control system sends the position of the front train to the rear train, and for the rear train, the position of the front train is a dangerous point, so that the speed limit of the rear train is calculated, the position of the front train is continuously changed, the speed limit of the rear train is changed accordingly, for the front train, the front station is a stopping point, the distance from the front train to the stopping point can be used for calculating the speed limit of the front train, a train target-distance mode curve can be obtained, the actual running speed of the train must strictly follow an ATP curve, and the speed limit cannot exceed the speed limit value, and fig. 2 is a train target-distance mode schematic diagram.

The train can be divided into four types of traction, cruising, coasting and braking according to the running state of the train. The traction is acceleration operation, the cruising is constant speed operation, the coasting is not to provide traction force to enable the train to slide by inertia, and the braking is deceleration operation. Defining resultant force C, tractive force F, resistance R and basic resistance R of train_bThe additional resistance being R_aAnd the braking force is B, then:

under the traction working condition, the resultant force exerted on the train is a positive value and is the same as the advancing direction, the train is in an accelerating state, and at the moment, the resultant force exerted on the train is as follows:

C＝F-R (11)

under the cruising condition, the resultant force borne by the train is 0, the train is in a constant speed state, the traction force is equal to the resistance, namely:

F＝R (12)

under the idle working condition, the traction force of the train is 0, the train slides forwards by means of inertia, the resultant force is equal to the resistance and is opposite to the advancing direction, namely:

C＝-R (13)

under the braking working condition, the resultant force borne by the train is opposite to the advancing direction, and the resultant force is in a deceleration state:

C＝-(B+R) (14)

the tractive force is not usually a fixed value, but changes with the change of the train speed, and the train traction characteristic curve is shown in fig. 3 by taking the CR400BF model as an example.

On the basis of a train dynamic model, the maximum allowable speed of a back train and a front train can be obtained according to the departure interval time of the train by combining a traction characteristic curve and other related parameters of the train, so that the speed-position curve of each train is determined, and the actual running time of the train is calculated.

(2) Power supply capability constraint

The power source required in the process of advancing the high-speed train is electric energy, the electric power is transmitted to the high-speed train through a contact network, when the train is close to each other, the possibility of insufficient power supply exists, and if a power supply system cannot meet the normal requirement of the train, the condition that the train is accelerated and decelerated inefficiently or forcibly occurs, so that the power supply capacity is another important constraint condition.

In a high-speed railway line in China, the voltage provided by a contact network is power frequency single-phase alternating current 25kV and 50Hz, and the actual voltage is 27.5kV considering the influence of impedance factors.

The actual magnitude of the traction current of the train in the running process is related to various factors such as the working condition, the running speed, the required traction force, the resistance and the like in the running process of the train, the traction current value at certain specific speed can be obtained according to data obtained by experiments through a large amount of theoretical calculation and data analysis of an actual line, and the current values at other speeds are determined in a linear fitting mode.

The schematic diagram of the traction power supply is shown below fig. 4:

as shown in fig. 4, there are two trains in the first power supply section, and three trains in the second power supply section, and the partial pressure of the train 1 is:

U₁＝U₀-(I₁+I₂)×s₁×R₀ (15)

U₀for drawing the wire, R₀The unit impedance is 200m omega/km, s₁Distance between train 1 and traction substation, I₁、I₂The traction currents of the train 1 and the train 2 are respectively.

The partial pressure of train 2 is:

U₂＝U₀-(I₁+I₂)×s₁×R₀-I₂×(s₂-s₁)×R₀ (16)

wherein s is₂The distance from the train 2 to the traction substation.

Similarly, the partial pressure of the train 3 is:

U₃＝U₀-(I₃+I₄+I₅)×s₃×R₀ (17)

the partial pressure of train 4 is:

U₄＝U₀-(I₃+I₄+I₅)×s₃×R₀-(I₄+I₅)×(s₄-s₃)×R₀ (18)

the partial pressure of the train 5 is:

U₅＝U₀-(I₃+I₄+I₅)×s₃×R₀-(I₄+I₅)×(s₄-s₃)×R₀-I₅×(s₅-s₄)×R₀ (19)

s₃、s₄、s₅the distances between the train 3, the train 4 and the train 5 and the traction substation, I₃、I₄、I₅The traction currents of the train 3, the train 4 and the train 5 are respectively.

According to the speed-position relation in the running process of the train, the traction current and the partial pressure of each train can be determined, and whether the adjustment scheme meets the constraint of the power supply capacity or not is further judged.

6. Genetic algorithm solving process for high-speed train operation adjustment problem

(1) Encoding

In the invention of the patent, the departure interval time of each train is described by taking minutes as a unit, and the interval time of each train can be used as a variable to participate in operation in a decimal mode. For example: and 6 trains are involved in adjustment, the solution of the problem can be represented in the form of 5 decimal numbers, for example, {3, 5, 2, 4, 3} is taken as a solution of the problem, and second, third, fourth, fifth and sixth trains are dispatched 3, 5, 2, 4 and 3 minutes after the first, second, third, fourth and fifth trains are dispatched respectively, and then the actual operation schedule is determined by the dispatching time of the first train, such as the first train 12: and 30, departure time of each train is { 12: 33, 12: 38, 12: 40, 12: 44, 12: 47, the problem can be described intuitively and effectively by adopting the coding mode, and the subsequent calculation is convenient.

(2) Objective function

The target function is the expression reflecting the high quality and the low quality of the solution of the problem, the invention takes the late time and the energy consumption as the optimization target, and the point to be noticed is that aiming at the characteristics and the practical application scene of the train operation adjustment problem, when the late time is smaller and the energy consumption is lower, the effect of the adjustment scheme corresponding to the obtained solution is better, namely the quality degree of the adjustment scheme is negatively related to the target function.

(3) Genetic parameters

In the genetic algorithm, three main genetic operations of selection, crossing and mutation are included, and through the operation, an individual can complete the optimal solution of the problems of excellence and disadvantage and gradual approximation. The selection means that the solution with higher quality directly enters the next generation by copying, the crossover finger exchanges partial genes of two chromosomes, the variation means that partial genes of one chromosome are randomly changed, and the crossover and variation are both ways of generating new individuals and searching the solution with higher quality. And each genetic operation needs to be completed according to a certain probability, the performance quality degree of the algorithm is generally greatly influenced by the probability, taking intersection as an example, if the probability is too small, the number of new individuals generated in each iteration is small, the population evolution speed is slow, the time required for solving is long, and if the probability is too large, the elite individuals in the population are easily damaged. Aiming at the characteristics of the algorithm and the principle of genetic operation, the following self-adaptive mode is designed in the algorithm to calculate the probability of cross and variation, the probability can be dynamically adjusted according to the quality degree of individuals in each generation of population by calculating according to the mode, the performance of the algorithm is improved, and the specific calculation mode is as follows:

in formulae (20) and (21), P_cRepresenting the cross probability, P_mRepresenting the probability of variation, k_cRepresenting the cross probability base value, k_mRepresenting the base value of the probability of variation, f_maxRepresents the maximum value of individual fitness in the population, f_cRepresenting the greater fitness of the two chromosomes undergoing crossover operation, f_mRepresenting the fitness value of the chromosome on which the mutation was performed,

represents the average of fitness of all chromosomes in the population.

According to the two formulas, the individual with higher adaptability is selected to have higher probability of crossing and mutation, so that the diversity of the population can be enhanced, the probability that the next generation population obtains more excellent individuals is higher, otherwise, the elite individual with lower adaptability is selected to have lower probability of crossing and mutation, and is usually directly stored by selection to enter the next generation population, and in sum, the three genetic operations can be enabled to promote the population to gradually develop towards the direction of higher solution quality after the corresponding probability is set.

(4) Termination conditions

In the process of solving the algorithm, each genetic operation can lead the population to be continuously evolved, the individual quality is higher and higher, but the algorithm cannot be operated infinitely, a termination condition needs to be set, whether the newly obtained population reaches the termination condition of the algorithm or not is judged for each generation of the newly obtained population, whether the calculation of the algorithm is continued or stopped is determined, and in the research of the algorithm, three termination conditions are designed. Reaching the maximum iteration times; the longest calculation time is exceeded; and thirdly, converging the population.

The method comprises the following steps of firstly, judging whether the algorithm is in an infinite operation mode or not, if the algorithm parameter is not reasonable in setting or data is wrong, searching the optimal solution, and if the algorithm reaches the maximum iteration times or exceeds the specified time, judging whether the program is in an abnormal operation mode or is wrong, searching the optimal solution, stopping operation, checking basic parameters such as the number of groups, the probability of crossing and variation, the weight of each index in an objective function, the number of trains and the like in the algorithm, calculating functions such as operation time and energy consumption, adjusting unreasonable parameters, correcting the calculated functions with mistakes, and calculating and solving again.

The judgment condition of the condition (c) is as follows:

in the formula (f)_iIs the fitness of the ith chromosome and is,

is population average fitness and is a small constant. If the population fitness satisfies the above formula, it can be considered that the population has reached convergence, the individuals in the population are the optimal solution, and the continuous operation will not changeThe original individual is changed, and at this time, the operation can be ended and the optimal solution obtained by calculation can be output.

(5) Genetic processes

The first generation population is randomly generated by a computer, the fitness distribution of each individual presents stronger uncertainty, and then the individual is selected, crossed and mutated on the basis of the first generation population, so that the population is continuously evolved and develops towards a better individual direction.

And (3) carrying out genetic operation according to the parameters, gradually eliminating poorer individuals, selecting excellent individuals, repeating the operation, and when the population reaches a stable state and the fitness is not reduced any more, determining the individuals in the population as the optimal individuals, and carrying out reverse operation according to the coding principle to complete the decoding process so as to obtain the optimal adjustment scheme.

The invention has the following beneficial effects: the method adopts a multi-constraint condition multi-optimization target genetic algorithm model, considers the problems of speed limit and power supply capacity in a train tracking operation scene, establishes an optimization target with minimum late time and energy consumption, and designs an optimal adjustment scheme calculation method based on a genetic algorithm to ensure that the line order is recovered as soon as possible. The adjusting scheme provided by the invention can shorten the time at the night as far as possible under the condition of ensuring the safe tracking distance of the train and the power supply, and has important practical significance.

In order to verify the effectiveness of the multi-constraint train operation plan solving method provided by the patent, the section completes software simulation on the optimal adjustment scheme for solving the genetic algorithm and analyzes the experimental result. Here, a three-station two-interval line from the Shanghai rainbow bridge to the Hangzhou east station at the middle warp stop Jiaxing south station is selected as a simulation example: the train number is G7375, G7595, G1389, G7323, G1353 and G1329 respectively, the length of the section from the Shanghai Rainbow bridge to the Jiaxing south and the Jiaxing south to the east station of Hangzhou is 90.6 kilometers and 79.5 kilometers respectively, the maximum speed limit of the whole-course line is 350km/h, and the basic parameters of the genetic algorithm are set as follows: the population number is 20, the chromosome length is 10, the first 5 bits represent departure interval time of Shanghai hong bridge station, and the last 5 bits represent departure of Jiaxing south stationThe interval time and the maximum iteration times are 15, the weight of the late time and the energy consumption in the objective function is 0.7 and 0.3 respectively, and in the normalization processing method of the objective function, the maximum late time and the minimum late time are determined in the following mode: if the train is dispatched according to the planned running time without adjustment, the total delay time is 180 minutes, and if the train is dispatched according to the shortest interval, the total delay time is 120 minutes, so that the maximum delay time and the minimum delay time can be respectively set to be 180 minutes and 120 minutes; the maximum and minimum energy consumption are determined as follows: if each train is not influenced by ATP speed limit, the train is accelerated to the maximum speed limit of the line by the maximum acceleration and then cruises to run, the running distance of the train under the traction working condition is 8.276km and the running distance under the cruises working condition is 53.524km through calculation, and the maximum energy consumption of the single train is 3.726 multiplied by 10 by combining the traction and resistance characteristics⁶kJ, total energy consumption of 2.236X 10⁷kJ, if each train runs at 1 minute intervals, the maximum running speed of the subsequent trains is reduced, and the total energy consumption is 1.573 multiplied by 10 through simulation calculation⁷kJ, the maximum and minimum total energy consumption are 2.236 × 10⁷kJ、1.573× 10⁷kJ。

The lengths of the power supply sections from shanghai rainbow bridge to jiaxing south station and from jiaxing south to hangzhou east station are shown in tables 1 and 2:

TABLE 1 Length of Power supply zone from Shanghai hong Qiao to Jiaxing Nan

Power supply section	1	2	3	4
					Length of	12km	23km	27km	29km

TABLE 2 length of Jiaxing south to Hangzhou east power supply region

Power supply section	1	2	3	4
					Length of	10km	20km	24km	26km

The train schedule of the above scenario is shown in table 3:

TABLE 3 planned hours of operation from Shanghai Rainbow bridge to Hangzhou east station

According to the data, setting a scene of a late point as a first train 30min late point caused by equipment failure, forcing a subsequent train to stop for waiting under the influence of the equipment failure, solving by adopting the genetic algorithm, wherein chromosomes randomly generated in a first generation population are integer arrays with the length of 10, the first train can run according to the maximum speed limit of a line because no other trains are influenced in front, then sequentially calculating the corresponding departure interval time after decoding the generated random arrays, calculating the maximum speed limit of the subsequent trains according to the positions of the previous trains, obtaining the speed-position curve of each train in the interval running process and the traction current and the partial pressure of each moment under the constraint, obtaining the interval running time length and the running distance of each train under the working conditions of traction and cruise from the running curve, and obtaining the actual arrival time of the next station by combining the total interval running time length and the departure time, and subtracting the planned arrival time to obtain the time of the late point, combining the running distance under the working conditions of traction and cruising with a traction characteristic curve and basic train parameters to obtain energy consumption, obtaining whether power supply capacity constraint is met or not through partial pressure, finally expressing two indexes of the total time of the late point and the total energy consumption as a target function after normalization processing, judging whether the power supply capacity constraint is met or not, if the power supply capacity constraint is met, taking a target function value as an adaptability value, if the power supply capacity constraint is not met, adding 1 to the target function value to obtain the adaptability value, wherein the first generation of population is shown in figure 5:

from the figure, 20 chromosomes generated by the first generation population have strong randomness and present the characteristic of random distribution corresponding to late time, energy consumption and fitness, except that the fitness values of the 4 th, 9 th, 13 th, 18 th and 20 th chromosomes are larger than 1 because the power supply capacity constraint is not met, the fitness values of the other individuals are different from about 0.568 to 0.930, the population is selected, crossed and mutated on the basis of the first generation population, and the individuals with higher quality are sought step by step.

The algorithm converges when the genetic manipulation proceeds to the seventh generation, which is shown in fig. 6, which is a seventh generation population diagram. All chromosomes in the population are {2, 5, 5, 4, 4, 3, 3, 2, 3, 4}, namely the departure time of the first station of 6 trains participating in adjustment is {13:21, 13:23, 13:28, 13:33, 13:37, 1 in the case of 30 minutes of the first train and the later point of the first train respectively3:41, the train stops for 2 minutes at the intermediate station, the arrival time of the second station is {13:49, 13:51, 13:56, 14:01, 14:05, 14:09} respectively, the departure time of the second station is {13:51, 13:54, 13:57, 13:59, 14:02, 14:06} respectively according to the algorithm result, but the adjustment according to the scheme can cause that part of the trains do not meet the stop constraint condition because the departure time is earlier than the arrival time, the final arrival time is {13:51, 13:54, 13:58, 14:03, 14:07, 14:11} and the final arrival time is {14:14, 14:17, 14:21, 14:26, 14:37, 14:40} respectively, and the final arrival time of each train is 30 late after the planned operation diagram, 18. 17, 11, 0 minutes, total late time to end station of about 93 minutes, total energy consumption of about 3.87 x 10⁷kJ, the fitness after standardization is about 0.5044, which is obviously reduced compared with the average fitness of 0.947 of the first generation population, and the fitness does not change in the subsequent inheritance, namely the fitness reaches the optimal fitness.

The calculation of the genetic algorithm is a process of continuously repeating genetic operation and gradually optimizing, the randomly generated individuals in the initial population have larger difference and more difference of fitness values, and in the subsequent generation of populations, the difference is gradually reduced, the quality of the individuals is gradually improved, and finally the individuals are converged to a stable value. To facilitate the description of the variation of the population of each generation, the average value of the two evaluation indexes and fitness of all individuals in the population can be used. Under the scene described herein, the two indexes of the total late point time average value and the total energy consumption average value of each generation of population in the genetic process are in a significantly decreased state, and finally reach a convergence state in the seventh generation, and are not decreased in the later genetics, as shown in fig. 7, which is a schematic diagram of the total late point time variation of the genetic process; and FIG. 8 is a schematic diagram of total energy consumption changes in a genetic process:

the change of the fitness function calculated by the standardization method is shown in fig. 9, a schematic diagram of the change of the fitness value in the genetic process is shown, the departure scheme of the fitness function is arranged for 6 times of trains at a later point according to the solution after convergence in the seventh generation of the algorithm, namely the adjustment scheme corresponding to the chromosome with the lowest fitness arranges the departure scheme of the train for 6 times at the later point, so that after the later point, the later point can be effectively relieved by shortening the departure interval, but under the constraint of ATP speed limit, the speed of subsequent trains is reduced due to too short departure interval time, the operation time is increased, the power supply capacity constraint cannot be met due to the fact that the number of trains in the same power supply section is large, the determination of the departure interval time needs to be solved through the algorithm, in the example, the genetic algorithm better solves the problem, an effective method is provided for the train operation adjustment problem, and therefore, the implementation process of the invention conforms, the final obtained result shows that the calculation effect is good and the practicability is high.

The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims (7)

1. A method for calculating multi-constraint train operation adjustment is characterized by comprising the following steps:

s1, modeling the high-speed train operation adjustment problem based on the genetic algorithm;

s2, constructing an optimization target comprehensively considering the late time and the energy consumption of the train according to the established model;

and S3, under the constraint condition, adjusting the train according to the optimization target.

2. The method of claim 1, wherein step 3 comprises:

s31, representing the train operation adjustment plan in the form of train operation interval time;

s32, the train runs at full speed according to the maximum speed limit of the line for the first time, the maximum speed limit is sequentially calculated according to the position of the train and the position of the preceding train determined by the interval time of each subsequent time, a speed-position curve is generated, and the interval running time is determined;

s33, determining the actual arrival time according to the departure interval time and the interval running time of each train;

s34, determining the late time by combining the plan operation chart;

s35, calculating energy consumption according to the speed-position curve;

s36, weighting and summing the late time and energy consumption to obtain a target function as a basis for determining the quality degree of the adjustment scheme;

s37, entering the optimizing process of the population according to the selection, crossing and variation of the genetic algorithm;

and S38, judging whether the algorithm reaches the convergence termination condition, if not, jumping to the step S31, otherwise, ending the processing, and taking the current result as the optimal adjustment scheme.

3. The method of claim 1, wherein step 3 comprises:

step 31, generating a first generation population by a random number generating mode through a computer, wherein population individuals are larger than a preset threshold value so as to ensure randomness and facilitate subsequent calculation;

step 32, among the trains participating in the adjustment, if no other train in front of the train which starts for the first time is a tracking point, taking a station which arrives in front as the tracking point, accelerating to the maximum speed limit of the line according to the maximum traction force, keeping cruising, and calculating a speed-position curve in a mode that the maximum braking force is decelerated to the station in front to stop, taking a corresponding time step to improve the calculation accuracy, and storing the speed, the position, the time, the traction current and the voltage value of the train at each moment;

step 33, calculating the maximum allowable speed limit at each moment by using the train closest to the front as a tracking point for each train in subsequent running, obtaining a speed-position curve according to the speed limit, and when the departure interval is small, the speed limit of the train in the rear running is easily influenced by the front train and reduced; the method comprises the steps that each train is dispatched according to dispatching interval time, the position of a front vehicle is obtained at each moment, the maximum speed limit is calculated, each train is guaranteed not to exceed the speed limit to run, and the speed, the position, the time and the traction current value of each train are recorded at each moment;

step 34, obtaining the total running time of each train according to the speed-position curve, and determining the transition point of the traction and cruising working conditions and the running distance under each working condition;

step 35, calculating the arrival time according to the interval running time of each train and the departure time:

step 36, the difference between the actual arrival time and the planned arrival time is the time of the train at the later point, the sum of the time of the train at the later point is the total time of the later point, the energy consumption is the integral of the traction force to the displacement under the traction and cruise working conditions, and the time of the later point and the energy consumption are normalized according to different weights and then summed to obtain the fitness value:

Ψ₁＝q_T*T_Delay+q_E*E (4)

step 37, obtaining the partial pressure of the contact network in the current power supply interval according to the traction current and displacement data in the running process of each train, judging whether the power supply constraint condition is met, and if the power supply constraint condition is not met, adding one to the fitness value to serve as the final fitness;

Ψ₂＝Ψ₁+sgn(U) (5)

and step 38, entering next generation operation and repeating the processes from step 32 to step 37.

4. The method of claim 3, wherein the fitness is calculated by:

(1) decoding the individuals to obtain the departure interval time of each train participating in adjustment;

(2) determining departure time of each train according to departure interval time by referring to a timetable;

(3) according to the ATP speed limit constraint condition, sequentially calculating the speed, the position, the traction current, the partial pressure and other data of each subsequent train in the running process according to the position of the front train;

(4) judging whether the electric energy meets the power supply capacity constraint or not according to the required electric energy of the high-speed train by the actual current and voltage value;

(5) determining the actual arrival time of each train according to the running time of each train;

(6) calculating the late time according to the formula (8) according to the planned arrival time and the actual arrival time,

is the actual arrival time of the ith train,

the planned arrival time of the ith train;

(7) calculating the total energy consumption of each train in a mode of integrating the traction force with the displacement according to the working condition in the running process according to the formula (9), wherein F_i，TraThe magnitude of the traction force of the ith train under the traction working condition, s_i，TraDistance traveled under traction condition of i-th train, F_i，CruThe magnitude of the traction force s under the cruising condition of the ith train_i，CruThe cruising working condition is the cruising working condition lower running distance of the ith train;

(8) normalizing the total late time and the total energy consumption to obtain a fitness function, calculating according to a formula (10), and q_TWeight in the objective function for the late time, q_EFor the weight occupied by energy consumption in the objective function, T_MinTo minimum time of night, T_MaxTo maximum evening time, E_MinTo minimize energy consumption, E_MaxIs the most importantHigh power consumption, Ψ is an fitness value, sgn (U) is a function that determines whether a power supply capability constraint is satisfied, U_min、U_maxRespectively obtaining a minimum value and a maximum value of voltage required by normal operation, if the minimum value and the maximum value are not satisfied, obtaining a value of M, if the minimum value and the maximum value are satisfied, obtaining a value of 0, wherein M is a proper positive number, and selecting M as 1 as the fitness function value is between 0 and 1;

5. the method of claim 4,

the specific steps of the ATP speed limit constraint include:

on the basis of a train dynamic model, a maximum allowable speed of a rear train is obtained according to the departure interval time of the train by combining a traction characteristic curve and other related parameters of the train, so that a speed-position curve of each train is determined, and the actual running time of the train is calculated;

the specific steps of the power supply capacity constraint judgment comprise:

and determining the traction current and the partial pressure of each train according to the speed-position relation in the running process of the trains, and further judging whether the adjustment scheme meets the constraint of the power supply capacity.

6. The method of claim 3,

the parameters of the genetic algorithm are determined according to: in the genetic algorithm, three main genetic operations of selection, crossing and variation are included, and through the operation, an individual can complete the optimal solution of the problem of the excellence and the disadvantage and gradually approaches; the selection means that the solution with higher quality directly enters the next generation by copying, the crossover finger exchanges partial genes of two chromosomes, the variation means that partial genes of one chromosome are randomly changed, and the crossover and variation are both ways of generating new individuals and searching the solution with higher quality. And each genetic operation needs to be completed according to a certain probability, and the specific calculation mode is as follows:

in formulae (20) and (21), P_cRepresenting the cross probability, P_mRepresenting the probability of variation, k_cRepresenting the cross probability base value, k_mRepresenting the base value of the probability of variation, f_maxRepresents the maximum value of individual fitness in the population, f_cRepresenting the greater fitness of the two chromosomes undergoing crossover operation, f_mRepresenting the fitness value of the chromosome on which the mutation was performed,

represents the average of fitness of all chromosomes in the population.

The probability of crossing and mutation of the selected individuals with high adaptability is higher, otherwise, the probability of crossing and mutation of the selected elite individuals with low adaptability is lower, and the elite individuals with low adaptability generally enter the next generation population directly through selection and are set with corresponding probabilities, so that the population is promoted to develop gradually towards the direction of higher solution quality by the three genetic operations.

The termination conditions include: reaching the maximum iteration times; the longest calculation time is exceeded; and thirdly, converging the population.

The judgment condition of the condition (c) is as follows:

in the formula (f)_iIs the fitness of the ith chromosome and is,

is population average fitness and is a small constant. If the population fitness satisfies the above formula, it can be considered that the population has reached convergence, the individuals in the population are already the optimal solution, the original individuals are not changed any more by continuing the operation, and at this time, the operation can be ended and the optimal solution obtained by the calculation can be output.

7. A multi-constraint train operation adjustment computing device, comprising:

the modeling unit is used for modeling the high-speed train operation adjustment problem based on the genetic algorithm;

the building unit is used for building an optimization target comprehensively considering the late time and the energy consumption of the train according to the built model;

and the adjusting unit adjusts the train according to the optimization target under the constraint condition.

Images