Working Condition Identification And Fuzzy Controller Optimization Of Composite Energy Storage Loader Based On BP Neural Network

In recent years,with the rapid development of China's economy year by year,the demand for infrastructure construction was gradually increasing,and the performance requirements of engineering vehicles were also getting higher and higher.However,due to the working characteristics such as low speed,heavy load and frequent starting and braking,the emission performance of engineering vehicles was barely satisfactory.Especially with the emergence of energy problems and the increasingly stringent emission requirements,it was an inevitable trend for the development of engineering vehicles to improve the economic performance.Hybrid power technology was widely used in ordinary passenger cars,with a good effect on energy saving and emission reduction.Therefore,taking the loader as an example,members of this project developed a hybrid power system with composite energy storage-engine,battery and hydraulic accumulator.The model-selection and parameter-matching of key components were conducted,and the key parameters of each subsystem were collaboratively optimized as well.However,the hybrid power system of composite energy storage loader was not able to choose the optimal control strategy according to the working condition,which caused that the control performance didn't reach the optimal state.Therefore,it was necessary to identify the working conditions based on the research of project members,and optimize the control strategy against different working conditions,so as to improve the economic performance of loaders as high as possible without loss of power.The specific contents of this paper were as follows:(1)By analyzing the general working mode and the power structure of the loader,a theoretically mathematical model was established as per the parameters.Based on this model,the backward simulation model of composite energy storage loader was built by using MATLAB/Simulink software.(2)By analyzing the main factors of the working condition,three initial decision vectors were obtained,and a BP neural network suitable for identifying the working conditions of loaders was created by using MATLAB software.The BP neural network was trained by selecting appropriate algorithms and training functions.(3)Targeting at the fuel consumption of the engine,with the help of the genetic algorithm,the membership function partition of the fuzzy control strategy of the loader's composite energy storage system under different working conditions was optimized.Therefore,the subjective blindness in the setting of membership function was eliminated and the parameter setting was more targeted.(4)The Simulink model of BP neural network generated through m file and the optimized fuzzy controller for different working conditions were integrated into the back-ward simulation model of the whole vehicle to conduct the simulation analysis on the control performance.The results showed that the BP neural network model was able to identify automatically working conditions,and the fuel economy of the whole vehicle was obviously improved after the optimization of the fuzzy controller.(5)Through hardware-in-the-loop tests on the dSPACE test bench in the laboratory,and comparing the test result with the simulation result,it was found that the two results were basically tallied with each other.This demonstrates that the working condition identification module based on the BP neural network and the optimization of the fuzzy controller based on the genetic algorithm is feasible and effective,which provides a reference for the design of the vehicle controller.