Parameters Optimization Design Of Vehicle Suspension Based On Intelligent Algorithm

Based on vehicle suspension system research status,and combination simulation and test analysis,the paper gave systematic research on ride comfort and road damage.By studying three side blotched lizard's behavior characteristics and survival mechanism,it put forward a tri-objective co-evolutionary algorithm based on lizard's behavior bionic.Combined with game theory,the bionic algorithm is used to vehicle ride comfort and handling stability.The paper also put forward a modified Elman Neural Network PAC controller to six-DOF suspension system test platform control.Through the experiment of parameters performance on wheel load utilization rate and frequency domain phase angle analysis,the paper provides a new research method for the suspension system design,and parameters selection.(1)Established eight DOF kinematic equations,A,B,C three kinds of road random excitation of road surface are simulated by white noise filtering method.According to vehicle ride comfort kinematics equation,the kinematics equations are solved by mode superposition method,which get the vehicle's modal characteristics and every DOF's displacement,velocity,acceleration vibration response,and analysis the stiffness and damping parameters of seat,suspension,wheel acting on vehicle performance.It also gives performance evaluation indexes of ride comfort,safety,space occupancy and respective evaluation content,which are: seat acceleration RMS,relative load RMS,suspension dynamic travel;andthe relevance are analyzed.(2)Competitive and cooperative game method is applied to optimize a linear vehicle suspension model with eight DOF.Considering road surface damage and ride comfort,taking stiffness and damping of seat and suspension system as design variables;and taking seat acceleration RMS,tire's relative dynamic load and suspension's maximum dynamic stroke as objective functions,it established a competitive and cooperative game design model and presented the optimal methods.By calculating the affecting factors of the design variables to objective functions and fuzzy clustering,the design variables are divided into different strategic spaces owned by each player.Based on competitive and cooperative game model,each game player takes its payoff as mono-objective to optimize its own strategic spaces and obtains the best strategy to deal with the others.All the best strategies are combined as a game strategy set.Compared with the initial design,results show that ride comfort,road surface damage and suspension dynamic travel are improved.Combined with MTS 320 4 channel tire coupling road simulator,vehicle frameunits,D2P(Development to Product)rapid prototype development platform and data acquisition systemintegration,it analysisand validated vehicle seat acceleration RMS,tire dynamic load RMS,suspension maximum dynamic process based on test.(3)According to the relationship of competitive and cooperative game and mechanism,the paper put forward a co-evolutionary algorithm for tri-objective optimization,which bionics on three male side blotched lizards specie'sbehavior characteristics and survival mechanism,depending on the orange throat male side blotched lizards as egoism.It looks orange throat male side blotched lizards as egoism,the blue throat male side blotched lizards as collectivism,Yellow throat male side blotched lizards as opportunism(self harm).And it takes three design objectives as three lizards and mapping design variables as lizard population's chromosome,three kinds of lizard's self-genetic factor are formed from the chromosomes.Based on three kinds of lizard's behavior,the mapping relationship between self-adaptive function and three objective functions has been established.A new chromosome has been made with the optimal genes.Based on converge condition,the optimal chromosome is obtained with multi-generation evolution.The cooperative co-evolutionary algorithm is used to solve the ride comfort and handling stability index,and the calculation results show the effectiveness and practicability of the proposed method.The tests proved that: the seat acceleration are decreased,and thefront?rear suspension dynamic travel RMS and dynamic tire load RMS have declined,the ride comfort and drivingsafety are improved.(4)In order to verify the validity of the proposed algorithm,the effect of suspension parameters in time domain and frequency domain on wheel load ratio and phase angle is studied.The system is composed of high frequency hydraulic vibration platform,suspension,damping spring,load mass,tire,hydraulic power station,data acquisition system,PAC controller,displacement sensor and acceleration sensor and so on.According to the basic Elman network ignores output layer nodes feedback,only meet the first-order linear dynamic signal processing system,which can't meet multi network and multi order system,the paper put forward an modified Elman neural network;Combined with the adaptive ant colony algorithm(ACO)information element evaporation factor?,as well as the size of heuristic search and characteristics to improve Elman network's weights and training,it realized intelligent searching and control.The improved and trained network enhances correlation layer and the feedback of output layer,and takes feedback gain as connection weight to train the network,which not only has time varying of proportion coefficient and integral coefficient,but also has strong adaptability,high learning efficiency and approximation accuracy.It established PAC controller based on modified Elman neural network,used to the six-DOF suspension system test platform control;it also analyzed the suspension damping,un-sprung mass,suspension stiffness,tire stiffness parameters to analysis the effect on utilization rate of tire-wheel load and phase angle,the paper provides a method for suspension parameter optimization design.