Tractor comprehensive performance optimization method based on genetic particle swarm hybrid algorithm

Abstract

The invention relates to a tractor comprehensive performance optimization method based on a genetic particle swarm hybrid algorithm. The method comprises the steps that traction performance, PTO performance, steering performance and fuel economy data of a tractor are collected in real time; collecting operation environment data; preprocessing the data; constructing a multi-objective optimization model, wherein objective functions of the multi-objective optimization model comprise operation efficiency, fuel economy, traction performance and steering performance; a genetic particle swarm hybrid algorithm is adopted to solve the model, and optimal operation parameters are generated; dynamically adjusting the operation mode of the tractor according to the optimal operation parameters; and performing historical data comparison and trend analysis by using a multi-objective optimization model, identifying a performance degradation trend, and providing maintenance suggestions. By combining the advantages of the genetic algorithm and the particle swarm optimization algorithm, the problems that the local search capability of the particle swarm optimization algorithm is poor and the convergence speed of the genetic algorithm is low are solved, the overall optimization of the comprehensive performance of the tractor is realized, the operation efficiency is improved, and the service life of equipment is prolonged.

Description

Technical Field

[0001] The present invention relates to the technical field of tractor control, and particularly relates to a method for optimizing the comprehensive performance of a tractor based on a genetic particle swarm hybrid algorithm. Background Art

[0002] Nowadays, the tractor industry has gradually comprehensive requirements for product functions. Tractors need to operate in different terrain environments such as hilly areas and plains, and the types of operations are complex and variable. The emphasis on the comprehensive performance of tractors varies in different working conditions, and different performance indicators will affect each other. For example, increasing the traction force will increase fuel consumption, which will lead to unreasonable distribution among the performances of tractors, increased power loss, reduced operation efficiency, and poor fuel economy.

[0003] To improve the operation efficiency of tractors and improve the utilization of tractor performance, it is necessary to collect detailed parameters of tractors through sensors, and obtain and optimize the relevant performances of tractors after data analysis. At present, the optimization methods proposed in China are to improve and analyze a certain performance index, and cannot optimize the comprehensive performance of tractors as a whole, and there are limitations in the results obtained from the analysis of tractor parameters. In addition, both the genetic algorithm or the particle swarm algorithm used in the existing methods have disadvantages. The genetic algorithm has strong global search ability but slow convergence speed; the particle swarm algorithm has fast convergence speed but is easy to fall into local optimal solutions. In the face of the multi-objective optimization task proposed by the present invention, these two algorithms cannot analyze and process multi-dimensional data in the task. Summary of the Invention

[0004] Aiming at the deficiencies of the above problems, the present invention proposes a method for optimizing the comprehensive performance of a tractor based on a genetic particle swarm hybrid algorithm to solve the problem that it is difficult for the optimization method based on a single algorithm in the prior art to process the multi-objective comprehensive data of tractors.

[0005] To solve the above problems, the technical solution adopted by the present invention is as follows: A method for optimizing the comprehensive performance of a tractor based on a genetic particle swarm hybrid algorithm, comprising:

[0006] The first step: Real-time collect the traction performance data, PTO performance data, steering performance data and fuel economy data of the tractor;

[0007] The second step: Collect the operation environment data, including soil humidity, slope and wind speed;

[0008] The third step: Preprocess the data collected in the first step and the second step, including filtering, missing value filling and normalization processing;

[0009] Step 4: Based on the preprocessed data, construct a multi-objective optimization model, where the objective functions of the multi-objective optimization model include operation efficiency, fuel economy, traction performance, and steering performance;

[0010] Step 5: Use a genetic particle swarm hybrid algorithm to solve the multi-objective optimization model and generate optimal operation parameters;

[0011] Step 6: Dynamically adjust the operation mode of the tractor according to the optimal operation parameters;

[0012] Step 7: Use the multi-objective optimization model for historical data comparison and trend analysis, identify the performance degradation trend, and provide maintenance suggestions.

[0013] As a preferred implementation manner, the objective function of the multi-objective optimization model is:

[0014] where x is the vector of tractor operation parameters to be optimized, f i (x) is the i-th sub-objective function, w i is the weight factor of the i-th sub-objective function, and m is the number of sub-objective functions; the sub-objective functions include operation efficiency, fuel economy, traction performance, and steering performance.

[0015] As a preferred implementation manner, the implementation of the genetic particle swarm hybrid algorithm in Step 5 includes the following steps:

[0016] S1: Initialize the population and randomly generate a set of solutions as the initial particle swarm;

[0017] S2: Obtain the fitness function based on the objective function of the multi-objective optimization model, and calculate the fitness value of each particle according to the fitness function;

[0018] S3: Update the particle swarm using the selection, crossover, and mutation operations of the genetic algorithm;

[0019] S4: Adjust the particle positions using the velocity and position update formulas of the particle swarm optimization algorithm;

[0020] S5: Iteratively optimize until the termination condition is met.

[0021] As a preferred implementation manner, the initialization of the population in Step S1 includes: determining the search space, including engine speed, PTO load distribution, steering sensitivity, and tire pressure; setting the population size, maximum number of iterations, convergence threshold, adjustment factor, and random factor; the adjustment factor is used to balance each objective function in the multi-objective optimization, and the random factor is used to increase the diversity of solutions.

[0022] As a preferred embodiment, the selection, crossover, and mutation operations in step S3 include: using roulette wheel selection or tournament selection methods to select excellent individuals according to fitness values; simulating chromosome exchange combinations and using single-point crossover or multi-point crossover methods to generate new excellent individuals; randomly selecting an individual from the population for mutation to increase the diversity of solutions.

[0023] As a preferred embodiment, the velocity and position update formulas of the particle swarm optimization algorithm in step S4 are as follows:

[0024] The velocity update formula is: v i (t + 1) = w·v i (t) + c1·r1·(pbest i -x i (t)) + c2·r2·(gbest - x i (t)) (2)

[0025] The position update formula is: x i (t + 1) = x i (t) + v i (t + 1) (3)

[0026] where w is the inertia weight, c1 and c2 are learning factors, r1 and r2 are random numbers, pbest i is the individual optimal solution, and gbest is the global optimal solution.

[0027] As a preferred embodiment, the termination conditions of the method include: reaching the maximum number of iterations, the objective function value reaching the convergence threshold, or the best fitness not improving for several consecutive generations.

[0028] As a preferred embodiment, the dynamic adjustment of the tractor operation mode in the sixth step includes: adjusting the engine speed, PTO load distribution, steering sensitivity, and tire pressure; combining the operation environment data to dynamically optimize the operation parameters.

[0029] As a preferred embodiment, the historical data comparison and trend analysis in the seventh step include: using time series analysis methods to predict the performance change trend; generating performance degradation warning information based on historical data.

[0030] As a preferred embodiment, the implementation of the method further includes: using MATLAB-Simulink co-simulation for integrated modeling, that is, writing a corresponding program in MATLAB to call Simulink to perform parametric modeling on the operating parameters of the tractor to obtain an integrated model of the tractor; feeding the optimal operating parameters into the integrated model for verification to ensure the feasibility and effectiveness of the optimization results.

[0031] Compared with the prior art, the optimization method for a single algorithm in the present invention has difficulty in dealing with the problem of multi-objective comprehensive data of tractors. Therefore, a comprehensive performance optimization method for tractors based on a genetic particle swarm hybrid algorithm is studied. First, by making the genetic algorithm (GA) and the particle swarm optimization (PSO) algorithm complement each other in advantages, the problems of poor local search ability of PSO and slow convergence speed of GA are solved. Second, compared with the existing single-performance analysis system, this method creates an integrated model to comprehensively analyze multiple performance functions of the tractor, is not limited to the influence of a certain type of parameter, and the evaluation results of the tractor are more comprehensive and objective. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 is a flowchart of a comprehensive performance optimization method for a tractor based on a genetic particle swarm hybrid algorithm according to the present invention;

[0033] Figure 2 is Figure 1 a flowchart of data preprocessing of the optimization method shown;

[0034] Figure 3 is Figure 1 a flowchart of the genetic particle swarm hybrid algorithm of the optimization method shown;

[0035] Figure 4 is Figure 1 a flowchart of dynamically adjusting the operating mode of the tractor of the optimization method shown;

[0036] Figure 5 is Figure 1 a flowchart of historical data analysis and early warning of the optimization method shown. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0037] The technical solutions of the present invention will be described clearly and completely below with reference to the accompanying drawings. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0038] As Figure 1 shown, a comprehensive performance optimization method for a tractor based on a genetic particle swarm hybrid algorithm provided by the present invention specifically includes the following steps:

[0039] Step 1: Collect real-time data on the traction performance, PTO (Power Take-Off) performance, steering performance, and fuel economy of the tractor.

[0040] Step 2: Collect data on the operating environment, including soil humidity, slope, and wind speed.

[0041] Specifically, sensors installed on the tractor are used to collect real-time data on the traction performance, PTO performance, steering performance, and fuel economy of the tractor. At the same time, data on the operating environment, including soil humidity, slope, and wind speed, are collected. The collection of these data is the basis for subsequent optimization, ensuring a comprehensive understanding of the tractor's performance in actual operations.

[0042] Step 3: Preprocess the data collected in Step 1 and Step 2, including filtering, missing value imputation, and normalization.

[0043] Specifically, as Figure 2 shown, the specific steps for preprocessing the data collected in Step 1 and Step 2 are as follows:

[0044] Filter the collected data to remove noise and ensure data accuracy.

[0045] Impute missing data using interpolation or other statistical methods to ensure data integrity.

[0046] Normalize the data to make it suitable for subsequent multi-objective optimization models. Among them, normalization is to ensure that data with different dimensions can be compared and optimized on the same scale.

[0047] This process can ensure data accuracy and consistency, providing high-quality data input for the construction of subsequent multi-objective optimization models.

[0048] Step 4: Based on the preprocessed data, construct a multi-objective optimization model. The objective functions of this model include operating efficiency, fuel economy, traction performance, and steering performance. Each objective function is assigned a corresponding weight to reflect the importance of different performance indicators in actual operations.

[0049] Among them, the objective functions of the multi-objective optimization model are:

[0050] Among them, x is the vector of tractor operating parameters to be optimized, f i (x) is the i-th sub-objective function, w i is the weight factor of the i-th sub-objective function, and m is the number of sub-objective functions; the sub-objective functions include operating efficiency, fuel economy, traction performance, and steering performance.

[0051] Step 5: Solve the multi-objective optimization model using a genetic particle swarm hybrid algorithm. This algorithm combines the global search ability of the genetic algorithm and the fast convergence characteristics of the particle swarm optimization algorithm, which can effectively avoid local optimal solutions and generate optimal operation parameters.

[0052] Specifically, as Figure 3 shown, the specific implementation steps of the genetic particle swarm hybrid algorithm are as follows:

[0053] S1. Initialize the population, randomly generate a set of solutions as the initial particle swarm. Among them, initializing the population includes determining the search space, including engine speed, PTO load distribution, steering sensitivity, and tire pressure; setting the population size, maximum number of iterations, convergence threshold, adjustment factor, and random factor; the adjustment factor is used to balance each objective function in multi-objective optimization, and the random factor is used to increase the diversity of solutions.

[0054] S2. Obtain the fitness function based on the objective function of the multi-objective optimization model, and calculate the fitness value of each particle according to the fitness function. Among them, the fitness function is based on the multi-objective optimization model, comprehensively considering multiple objective functions such as operation efficiency, fuel economy, traction performance, and steering performance.

[0055] Specifically, first, obtain the fitness function based on the objective function of the multi-objective optimization model. Secondly, determine whether the fitness value of each particle calculated according to the fitness function meets the termination condition. If it meets the termination condition, output the optimal solution; if not, proceed to step S3.

[0056] S3. Update the particle swarm using the selection, crossover, and mutation operations of the genetic algorithm. Among them, the selection operation adopts the roulette wheel or tournament selection method, the crossover operation adopts single-point or multi-point crossover, and the mutation operation randomly selects individuals for mutation to increase the diversity of solutions.

[0057] S4. Adjust the particle position using the velocity and position update formulas of the particle swarm optimization algorithm.

[0058] Among them, the velocity update formula is: v i (t + 1) = w·v i (t) + c1·r1·(pbest i - x i (t)) + c2·r2·(gbest - x i (t)) (2)

[0059] The position update formula is: x i (t + 1) = x i (t) + v i(t + 1)(3)

[0060] where w is the inertia weight, c1 and c2 are learning factors, r1 and r2 are random numbers, pbest i is the individual optimal solution, and gbest is the global optimal solution.

[0061] S5. Recursively calculate the fitness value, update the genetic algorithm, and adjust the particle swarm optimization until the termination condition is met. The termination conditions include reaching the maximum number of iterations, the objective function value reaching the convergence threshold, or the best fitness not improving for several consecutive generations.

[0062] Step 6: Dynamically adjust the operation mode of the tractor according to the optimal operation parameters generated by the multi-objective optimization model. Adjusting the operation mode of the tractor includes adjusting the engine speed, PTO load distribution, steering sensitivity, and tire pressure.

[0063] Specifically, as Figure 4 shown, dynamically adjust the operation modes such as engine speed, PTO load distribution, steering sensitivity, and tire pressure according to the optimal operation parameters generated by the multi-objective optimization model, and combine the operation environment data to optimize the operation parameters. This process combines the real-time collected operation environment data to ensure that the tractor can maintain the best performance under different working conditions.

[0064] Step 7: Use the multi-objective optimization model to conduct historical data comparison and trend analysis, identify the performance degradation trend, and provide maintenance suggestions.

[0065] Specifically, as Figure 5 shown, the specific steps of using historical data comparison and trend analysis are as follows:

[0066] Store the operation parameters and performance data after each optimization as historical data for subsequent analysis.

[0067] Use time series analysis methods such as autoregressive integrated moving average model (ARIMA) and time series prediction model (Prophet) to analyze the historical data and predict the performance change trend.

[0068] According to the performance change trend, identify the performance degradation trend and generate warning information. If the performance index continues to decline, the system will issue a warning to remind the user to perform maintenance.

[0069] Based on the performance degradation trend, the system generates maintenance suggestions to help the user take measures in advance and extend the equipment life.

[0070] As a preferred embodiment, the implementation of the method further includes: using MATLAB-Simulink co-simulation for integrated modeling, that is, writing corresponding programs in MATLAB to call Simulink to perform parametric modeling on the operating parameters of the tractor to obtain an integrated model of the tractor; feeding the optimal operating parameters into the integrated model for verification to ensure the feasibility and effectiveness of the optimization results.

[0071] In the present invention, by complementing the advantages of the genetic algorithm (GA) and the particle swarm optimization (PSO) algorithm, the problems of poor local search ability of PSO and slow convergence speed of GA are solved. Secondly, compared with the existing single-performance analysis system, this method creates a comprehensive model to comprehensively analyze multiple performance functions of the tractor, is not limited to the influence of a certain type of parameter, and the evaluation results of the tractor are more comprehensive and objective. Moreover, it can effectively optimize the comprehensive performance of the tractor, improve the operation efficiency, reduce the fuel consumption, and ensure the long-term stable operation of the equipment.

[0072] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A tractor comprehensive performance optimization method based on a genetic particle swarm hybrid algorithm, characterized in that It includes the following steps: Step 1: Collect the traction performance data, PTO performance data, steering performance data, and fuel economy data of the tractor in real time; Step 2: Collect the operation environment data, including soil humidity, slope, and wind speed; Step 3: Preprocess the data collected in Step 1 and Step 2, including filtering, missing value filling, and normalization processing; Step 4: Build a multi-objective optimization model based on the preprocessed data. The objective function of the multi-objective optimization model includes operation efficiency, fuel economy, traction performance, and steering performance; Step 5: Use a genetic particle swarm hybrid algorithm to solve the multi-objective optimization model and generate the optimal operation parameters; Step 6: Dynamically adjust the operation mode of the tractor according to the optimal operation parameters; Step 7: Use the multi-objective optimization model for historical data comparison and trend analysis, identify the performance degradation trend, and provide maintenance suggestions.

2. The optimization method for the comprehensive performance of a tractor based on a genetic particle swarm hybrid algorithm according to claim 1, wherein The objective function of the multi-objective optimization model is: where x is the vector of tractor operation parameters to be optimized, and f i (x) is the i-th sub-objective function, and w i is the weight factor of the i-th sub-objective function, and m is the number of sub-objective functions; the sub-objective functions include operation efficiency, fuel economy, traction performance, and steering performance.

3. The tractor comprehensive performance optimization method based on the genetic particle swarm hybrid algorithm according to claim 1, characterized in that The implementation of the genetic particle swarm hybrid algorithm in Step 5 includes the following steps: S1: Initialize the population and randomly generate a set of solutions as the initial particle swarm; S2: Obtain the fitness function based on the objective function of the multi-objective optimization model, and calculate the fitness value of each particle according to the fitness function; S3: Update the particle swarm using the selection, crossover, and mutation operations of the genetic algorithm; S4: Adjust the particle positions using the velocity and position update formulas of the particle swarm optimization algorithm; S5: Iteratively optimize until the termination condition is met.

4. The optimization method for the comprehensive performance of a tractor based on a genetic particle swarm hybrid algorithm according to claim 3, wherein The initialization of the population in Step S1 includes: determining the search space, including engine speed, PTO load distribution, steering sensitivity, and tire pressure; setting the population size, maximum number of iterations, convergence threshold, adjustment factor, and random factor; the adjustment factor is used to balance each objective function in the multi-objective optimization, and the random factor is used to increase the diversity of solutions.

5. The optimization method for the comprehensive performance of a tractor based on a genetic particle swarm hybrid algorithm according to claim 3, wherein The selection, crossover, and mutation operations in Step S3 include: adopting the roulette wheel selection or tournament selection method to select excellent individuals according to the fitness value; simulating chromosome exchange combinations and using the single-point crossover or multi-point crossover method to generate new excellent individuals; randomly selecting an individual from the population for mutation to increase the diversity of solutions.

6. The optimized method for the comprehensive performance of a tractor based on a genetic particle swarm hybrid algorithm according to claim 3, wherein The velocity and position update formulas of the particle swarm optimization algorithm in Step S4 are: The velocity update formula is: v i (t + 1)= w·v i (t)+ c1·r1·(pbest i - x i (t))+ c2·r2·(gbest - x i (t)) (2) The position update formula is: x i (t + 1)=x i (t)+v i (t + 1)(3) where w is the inertia weight, c1 and c2 are learning factors, r1 and r2 are random numbers, and pbest i is the individual optimal solution, and gbest is the global optimal solution.

7. The tractor comprehensive performance optimization method based on the genetic particle swarm hybrid algorithm according to claim 3, characterized in that, The termination conditions include: reaching the maximum number of iterations, the objective function value reaching the convergence threshold, or the best fitness not improving for several consecutive generations.

8. The optimization method for the comprehensive performance of a tractor based on a genetic particle swarm hybrid algorithm according to claim 1, wherein The dynamic adjustment of the operation mode of the tractor in Step 6 includes: adjusting the engine speed, PTO load distribution, steering sensitivity, and tire pressure; dynamically optimizing the operation parameters in combination with the operation environment data.

9. The optimized method for the comprehensive performance of a tractor based on the genetic particle swarm hybrid algorithm according to claim 1, characterized in that, The historical data comparison and trend analysis in Step 7 include: using the time series analysis method to predict the performance change trend; generating performance degradation warning information based on the historical data.

10. The tractor comprehensive performance optimization method based on the genetic particle swarm hybrid algorithm according to claim 1, characterized in that The method further includes: writing a corresponding program in MATLAB to call Simulink to perform parametric modeling on the operating parameters of the tractor to obtain an integrated model of the tractor; feeding the optimal operating parameters into the integrated model for verification to ensure the feasibility and effectiveness of the optimization results.

Images