Nonlinear Parameter Identification Of Hydro-Pneumatic Suspension And Its Semi-Active Nonlinear Control With Constraints

With the development of electronic control technology in modern vehicles,the semi-active suspension,which balanced the cost and performance,has broken the inherent limitations of the traditional suspension system and has been widely used in vehicles.The hydro-pneumatic suspension has relatively complex nonlinear mechanical behavior.In the past,its research mainly focused on the characteristics of hydro-pneumatic suspension,and its semi-active control was also dominated by linear control.To improve its control performance,it is urgent to propose an effective nonlinear control method combined with the characteristics of hydro-pneumatic suspension.At the same time,compared with the simulation model,the actual vehicle model will be affected by various factors,such as parameter uncertainty,sensor noise,inaccurate estimation from state observers,delays of actuators,etc.As a result,there is a large error between the actual and simulated control results,and it is even difficult to use the designed controller in practice.To fully simulate the influence of these factors and maximally reduce their influence on the control performance are key issues in the control field.In particular,these issues are more prominent for the off-highway vehicles equipped with hydropneumatic suspension with the highly complex nonlinear characteristics.For that,this thesis,driven by the national key research and development plan and the project commissioned by the enterprise,is committed to developing a semi-active control system for the mining dump truck with the hydropneumatic suspension.The identification method and stateless control method for a constrained semiactive hydro-pneumatic suspension system are proposed to achieve the best parameter matching in vehicle performance,which provides a theoretical basis for the nonlinear parameter identification and the design of the nonlinear controller for the nonlinear suspension system.The main research contents and conclusions are as follows.(1)The nonlinear parameter identification of hydro-pneumatic suspension is carried out by extracting the features in both time and frequency domains.The time-domain modeling method for random road based on inverse fast Fourier transformation(IFFT)method and the data processing method based on wavelet analysis are studied after the vehicle dynamics model with nonlinear hydropneumatic suspension is established.The optimization model for nonlinear parameter identification is constructed by considering the features in both time and frequency domains and solved by the intelligent simulated annealing algorithm.The nonlinear parameter identification of hydro-pneumatic suspension considering the features in both time and frequency domains is proposed by the series of integrations.The good accuracy and reliability of the identification are verified by the quantitative and qualitative analysis,and an effective solution is formed for the nonlinear parameter identification of hydro-pneumatic suspension or other nonlinear vibration systems.(2)Nonlinear control method for the semi-active system with the hydro-pneumatic suspension is presented.The dynamical equation with a minimized nonlinear term is constructed by combining the linear equivalent stiffness and damping,after the vehicle dynamic model of the semi-active hydropneumatic suspension is established.On this basis,an optimal model with the actuator constraint for minimizing the error between the actual and ideal overall control forces is established,which realizes the optimal approximation from the nonlinear control to the linear quadratic optimal control.The comparative results of the different control methods show the superiority of the nonlinear control method proposed in this thesis.It effectively combines the unconstrained nonlinear control with the constrained actuator,having a good comprehensive performance control effect.(3)The design and analysis of the state observer are carried out.The calculation methods of the Luenberger state observer and the Kalman filter observer are given,the estimated results of the two observers under three working conditions are compared,and their respective advantages and differences in estimation accuracy are shown.After considering the estimation accuracy,numerical stability,and simulation efficiency,the Luenberger state observer is selected for the estimation of the system state to design the controller.It promotes the organic combination between the state observer and the nonlinear control based on the averaging technology,which lays the foundation of simpler and practical control laws.(4)A nonlinear stateless control method for semi-active hydro-pneumatic suspension is proposed.To reduce the parameters of the control system,reduce the error propagation between modules,and improve the reliability of the control system,the derivation process of the organic combination of the controller and the state observer is studied,and a control law in which there is no intermediate state is obtained.The control law directly establishes the relationship between the control input and the historical input and output,and a nonlinear stateless control method for the semi-active hydropneumatic suspension is formed by solving the optimization model minimizing the error between the actual and ideal overall control forces.The comparative results of the different control methods show that:(i)the nonlinear stateless control largely overcomes the inaccurate estimation of the state estimator;(ii)in addition to the suspension dynamic stroke,the nonlinear stateless control has a good performance better than the skyhook control and passive control;(iii)compared with other control methods,the nonlinear stateless control whose response is smoother,and its comprehensive performance is the best.In addition,the evolutionary analysis of the nonlinear stateless control with the constraint for a semi-active system shows that the proposed method can be degenerated into a constrained nonlinear control problem by appropriately adjusting the statistical length,which shows that it has downward compatibility.It has strong applicability and greatly expands the scope of use of the controller.(5)The verification of the hydro-pneumatic suspension model,the reliability of the designed controller,and the bench experiment for the semi-active control system are given.The accuracy of the mathematical model for hydro-pneumatic suspension shown in this thesis is verified by the comparisons with the experimental results from the authoritative literature.According to the reliability theory and Monte Carlo sampling,the reliability of the nonlinear stateless control for the semi-active suspension under the uncertain parameters is analyzed by considering the performance index.A bench experiment for the semi-active suspension system is developed,which lays a foundation for further research on the actual control effect.By the studies of the modelling,simulation,parameter identification,and control for the nonlinear hydro-pneumatic suspension,a nonlinear parameter identification method with both timefrequency domain characteristics extraction is proposed.A nonlinear stateless control method with constraints for semi-active hydro-pneumatic suspension system were developed and the reliability analysis method of nonlinear control simulation under parameter uncertainty is formed.They provide an effective evaluation method for reliability analysis of controllable suspension system under parameter uncertainty,and also provide a theoretical basis for nonlinear parameter identification and nonlinear controller design.