WO2020215659A1 - 一种电液智能转向系统性能监测器及性能优化方法 - Google Patents

一种电液智能转向系统性能监测器及性能优化方法 Download PDF

Info

Publication number
WO2020215659A1
WO2020215659A1 PCT/CN2019/116038 CN2019116038W WO2020215659A1 WO 2020215659 A1 WO2020215659 A1 WO 2020215659A1 CN 2019116038 W CN2019116038 W CN 2019116038W WO 2020215659 A1 WO2020215659 A1 WO 2020215659A1
Authority
WO
WIPO (PCT)
Prior art keywords
steering
module
vehicle
signal
coefficient
Prior art date
Application number
PCT/CN2019/116038
Other languages
English (en)
French (fr)
Inventor
赵万忠
栾众楷
王春燕
周小川
Original Assignee
南京航空航天大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 南京航空航天大学 filed Critical 南京航空航天大学
Publication of WO2020215659A1 publication Critical patent/WO2020215659A1/zh

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B62—LAND VEHICLES FOR TRAVELLING OTHERWISE THAN ON RAILS
    • B62D—MOTOR VEHICLES; TRAILERS
    • B62D15/00—Steering not otherwise provided for
    • B62D15/02—Steering position indicators ; Steering position determination; Steering aids
    • B62D15/025—Active steering aids, e.g. helping the driver by actively influencing the steering system after environment evaluation
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B62—LAND VEHICLES FOR TRAVELLING OTHERWISE THAN ON RAILS
    • B62D—MOTOR VEHICLES; TRAILERS
    • B62D6/00—Arrangements for automatically controlling steering depending on driving conditions sensed and responded to, e.g. control circuits

Definitions

  • the invention belongs to the technical field of automobile power steering system control, and specifically refers to an electro-hydraulic intelligent steering system performance monitor and a performance optimization method.
  • the electro-hydraulic intelligent steering system is a new type of steering system for large and medium-sized commercial vehicles. It is characterized by the ability to switch the steering mode by coordinating the output ratio of the dual actuators, that is, the composite steering with large torque output at low vehicle speeds Mode, adopting a good road feeling and more energy-saving electric power steering mode at high speed is an ideal steering system design form.
  • the different parameter values in the electro-hydraulic intelligent steering system have a great influence on the performance of the system. Therefore, it is necessary to analyze the dynamic model of the steering system to obtain the quantitative formula of the performance index including its key structural parameters, so as to be able to Improve the performance of the steering system.
  • For the performance optimization design of the new electro-hydraulic hybrid steering system there are two problems: one is how to put forward the evaluation index and constraint conditions of the system to be optimized according to the design characteristics and dynamic characteristics of the model to be solved, so as to carry out the multi-objective optimization model Construction; Second, according to the characteristics of the constructed problem, design the corresponding numerical optimization algorithm to calculate the proposed problem.
  • a multi-objective optimization model is established for the target of displacement and wheel swing angle, and the preference level of each optimization target of the established mathematical model is constructed through the physical programming method.
  • the solution after the cycle and others quantified the electro-hydraulic power steering system damping and assist coefficients, and proposed an optimization model with the optimal steering wheel feeling as the objective function.
  • the above-mentioned studies only quantify performance indicators for the traditional vehicle handling and stability requirements, and do not take into account the current energy-saving requirements of electric vehicles and the performance requirements of electro-hydraulic intelligent steering systems, and propose corresponding evaluation indicators.
  • the weighted sum method is a commonly used multi-objective optimization problem solving method. Its principle is to integrate multiple objectives into one objective through weight factors and scale factors, that is, to convert the multi-objective optimization problem into one objective.
  • the single-objective optimization problem is calculated by the single-objective optimization method. This method is simple and practical in engineering and can directly obtain an optimal solution. However, if one of the objectives is too sensitive, the optimization results of other objectives may be poor.
  • the characteristic of the multi-objective evolutionary algorithm is that it uses "population" as the evolution carrier to realize parallel computing.
  • the algorithm is based on "fitness” to reflect the pros and cons of the new solution, and through crossover, mutation and other update strategies, so as to ensure the diversity of the new population and prevent the solution process from falling into a local optimum.
  • the commonly used multi-objective evolutionary algorithms are mainly designed based on the dominance strategy, which is characterized by the introduction of the concept of dominance and dominance, but it has the problem of too many non-dominated solutions in the later stage of the solution process, which affects the efficiency of the solution.
  • the solution principle of the MOEA/D algorithm designed based on the decomposition strategy is to decompose a multi-objective optimization (MOP) problem into multiple units through an "aggregate function".
  • MOP multi-objective optimization
  • SOP single-objective optimization
  • the evolutionary algorithm is used to simultaneously solve all SOP sub-problems based on the decomposition strategy in one iteration, and the weight vector with good characteristics is distributed in the target space to ensure the entire solution process.
  • population diversity that is, through the "neighborhood", the individuals in the population co-evolve with each other in a "small group” manner.
  • this strategy not only reduces the time complexity of the algorithm, but also improves the accuracy of understanding, but the existing fixed search strategy cannot achieve adaptive adjustment, so there is an efficiency problem in solving actual engineering optimization problems.
  • the MOEA/D multi-objective optimization algorithm architecture based on the decomposition strategy is a multi-objective problem solving method with low computational complexity and high accuracy.
  • the swarm intelligence algorithm represented by the PSO operator is robust It has the characteristics of strong performance, adaptability and rapid response to environmental changes. It is suitable for the calculation of engineering optimization problems. It is an ideal steering system performance dynamic optimization method and has broad application prospects.
  • the purpose of the present invention is to provide an electro-hydraulic intelligent steering system performance monitor and performance optimization method, so as to solve the inability of the prior art to perform real-time online monitoring and dynamic optimization of the current steering system performance. Calculate the problem.
  • An electro-hydraulic intelligent steering system performance monitor of the present invention includes: an information unit, a signal processing unit, a steering performance monitor unit, an execution unit, a steering transmission unit and a sensing unit;
  • the information unit includes a driver behavior information module and a vehicle environment information module;
  • the signal processing unit includes a sensor filtering module, an information fusion module and a state estimation module;
  • the steering performance monitor unit includes a steering road feel evaluation subsystem, a steering sensitivity evaluation subsystem, a steering energy consumption evaluation subsystem, and an online optimization solution module;
  • the execution unit includes an electric actuator and a hydraulic actuator
  • the steering transmission unit includes a steering gear module and a steering column module;
  • the sensing unit includes a dynamic parameter sensor and an environment sensing sensor; wherein,
  • the driver transmits the angle/torque signal a to the information unit through the steering wheel.
  • the driver behavior information module and the vehicle environment information module in the information unit respectively output the extracted driver input behavior signal b and surrounding vehicle and environment perception signals c To the signal processing unit; the sensor filter module filters the input signal, the information fusion module fuses the filtered signal, the state estimation module estimates the required unmeasured state variables, and outputs the fused and filtered signal d to the steering Performance monitor unit; the steering road feel evaluation subsystem, steering sensitivity evaluation subsystem, and steering energy consumption evaluation subsystem in the steering performance monitor unit are respectively responsible for steering road feel (the road information felt by the driver) and steering sensitivity (steering wheel).
  • the angle change causes the vehicle yaw rate change) and steering energy consumption to quantify the three steering system evaluation indicators
  • the online optimization solution module quantifies the steering road feel evaluation subsystem, steering sensitivity evaluation subsystem, and steering energy consumption evaluation subsystem.
  • the dynamic optimization calculation is performed to obtain the optimized control signal e and output to the execution unit; the electric actuator and hydraulic actuator in the execution unit output the steering torque f to the steering transmission unit and the steering column according to the optimized control signal e.
  • the module and the steering module sequentially transmit the steering torque f to realize the steering operation; the sensing unit collects the current generated by the steering transmission unit and the rotation angle signal g, and transmits the processed multi-sensor parameter signal h to the information unit , Realize the steering operation of the closed loop system.
  • the multi-sensor parameter signal includes: the yaw rate signal, the lateral acceleration signal, the pitch angle signal, the vehicle speed signal collected by the dynamic parameter sensor, and the distance signal between the host vehicle and the preceding vehicle collected by the environmental sensor, Signal for the relative speed of the main vehicle and the preceding vehicle, signal for the distance between the vehicle in the left lane and the main vehicle, signal for the relative speed of the vehicle in the left lane to the main vehicle, signal for the distance between the vehicle in the right lane and the main vehicle, and signal for the relative speed of the vehicle in the right lane to the main vehicle .
  • performance indicators of the steering road feel evaluation subsystem are as follows:
  • ⁇ 0 is the cut-off frequency
  • n e1 is the proportional coefficient between the steering screw angle and the hydraulic actuator motor angle
  • n e2 is the proportional coefficient between the steering screw angle and the electric actuator motor angle
  • l is the center distance of the steering screw
  • a P is the hydraulic cylinder area
  • q is the displacement of the vane pump
  • i g is the transmission ratio
  • ⁇ is the density of the hydraulic oil
  • P is the pitch
  • N is the pump speed
  • C q is the flow coefficient
  • a 1 is the pipe area (assuming the same four pipe cross-section)
  • K S is the stiffness sensor
  • K is the current coefficient
  • n m1 is the motor speed of the hydraulic actuator
  • n m2 for the electric actuator Motor speed
  • n 2 is the reduction ratio of the worm gear of the electric actuator
  • r w is the radius of the pinion gear
  • ⁇ is the speed of the vane pump
  • the performance indicators of the steering sensitivity evaluation subsystem are as follows:
  • m is the mass of the vehicle
  • m s is the sprung mass
  • u is the longitudinal speed
  • h is the height of the center of mass of the vehicle
  • I x is the moment of inertia around the x-axis of the vehicle
  • I z is the moment of inertia around the z-axis of the vehicle
  • I xz is the moment of inertia around the xz plane of the vehicle
  • L p is the equivalent spring mass coefficient
  • L ⁇ is the equivalent pitch angle coefficient
  • N ⁇ is the wheelbase front wheel rotation angle coefficient
  • N ⁇ is the wheelbase roll angle coefficient
  • N ⁇ is Wheelbase side slip angle coefficient
  • N r is the wheelbase yaw rate coefficient
  • Y r is the suspension yaw rate coefficient
  • Y ⁇ is the suspension front wheel rotation angle coefficient
  • Y ⁇ is the suspension pitch angle coefficient
  • Y ⁇ is the suspension Frame slip angle coefficient
  • a is the distance from the center of
  • performance indicators of the steering energy consumption evaluation subsystem are as follows:
  • the total energy consumption equation of the electro-hydraulic compound steering system is shown in the following formula.
  • the main power consumption of the system includes four parts: controller power consumption P 1 , motor power consumption P 2 , hydraulic pump power consumption P 3 and rotary valve power consumption P 4 ;
  • E 1 P 1 +P 2 +P 3 +P 4
  • the controller power consumption P 1 is shown in the formula:
  • R A is the armature resistance
  • I A is the armature current
  • U c is the controller voltage
  • R elec is the controller resistance
  • p is the loss during transmission
  • f is the changing frequency of the alternating magnetic field
  • B m is the magnetic density amplitude
  • ⁇ is the Steinmetz coefficient
  • k h , k c and k e are the hysteresis loss coefficient, classical eddy current loss coefficient and eddy current loss coefficient, respectively ;
  • ⁇ is the motor speed
  • Q s is the pump flow
  • P s is the inlet pressure of the power cylinder
  • q is the oil pump displacement
  • ⁇ is the oil density
  • C q is the flow coefficient
  • a 1 and A 2 are the throttling areas of valve ports 1 , 2 respectively
  • Q s is the flow rate of the hydraulic pump
  • Ap is the effective area of the hydraulic cylinder piston
  • x r is The displacement of the steering nut.
  • An electro-hydraulic intelligent steering system performance optimization method of the present invention includes the following steps:
  • Step 1 Problem definition, including model definition and algorithm parameter definition
  • Step 2 Initialization: Iterate each particle, assign particle position and velocity, and calculate fitness function until the end of the loop;
  • Step 3 The main loop: The main loop of the algorithm is composed of the basic particle swarm optimization (PSO) module, the adaptive decomposition operator (AD) decomposition module and the Pareto module; first, the particle velocity, position and fitness function values are determined by the basic particles The swarm algorithm module is updated; secondly, the adaptive decomposition operator decomposition module is used to decompose and search the particles updated by the basic particle swarm algorithm module; finally, the loop termination condition is judged, and the Pareto set is derived;
  • PSO basic particle swarm optimization
  • AD adaptive decomposition operator
  • Step 4 If the end condition is not met, the search process returns to step 3, otherwise the Pareto solution set is derived.
  • step 1 specifically includes:
  • Model definition includes: model definition, optimization goals, constraints and design variables;
  • the algorithm parameter definition as compared with MOPSO, not only the maximum number of iterations defined algorithm I te, the number of particles n ori, inertia weight w, the weight decrease rate w damp, individual learning factor c 1, global learning factor c 2, Pareto set
  • the threshold n TPareto also needs to determine the rank of the neighborhood of the decomposition module.
  • step 3 specifically includes:
  • v i,j (t+1) and x i,j (t+1) are the velocity and position of the particle at t+1, respectively, v i,j (t), x i,j (t) Are the speed and position of the particle at time t, c 1 is the individual learning factor, r 1 is the individual learning factor weight, c 2 is the global learning factor, r 2 is the global learning factor weight, p i, j are the current search process , P g,j is the current global optimal particle;
  • AD Adaptive decomposition operator
  • the decomposition search strategy is divided into two parts. First, the neighbor of the current sub-problem is determined by the distance matrix to operate the decomposition search; then, an adaptive search is performed in each search direction determined by the weight matrix.
  • the adaptive search process mainly includes: Establish a multi-objective optimization model with d 1 (p) and d 2 (p) as optimization goals and algorithm robustness criteria as constraints. By adjusting the design variable p, when the optimization model meets the robustness constraints, d 1 ( p) and d 2 (p) tend to be the smallest; the optimization model is shown in the following formula:
  • d 1 (p) is the particle convergence distance
  • d 2 (p) is the particle diversity distance
  • p is the design variable of the decomposition search module
  • w, p) is the search method
  • x* is the ideal reference point
  • x k is the point obtained by the actual search
  • w is the weight coefficient of the decomposition search module
  • x and y be the two solution points obtained after one iteration, let x.object(j) denote the fitness function value of object j corresponding to particle x; for the multi-objective optimization problem with the smallest objective value, if all x.object (j) is less than or equal to y.object(j), and at least one x.object(j) is less than y.object(j), then x belongs to the Pareto solution set.
  • the present invention processes and analyzes the collected vehicle signals, environmental signals, and driver signals through a sensor filter module, an information fusion module, and a state estimation module, and then transmits them to a performance monitor.
  • the performance monitor includes steering feel, steering sensitivity, and
  • the three evaluation subsystems of steering energy consumption can optimize and solve the current system performance online according to the evaluation equation to obtain the best steering behavior decision, and then input the optimized control signal to the lower execution unit to realize the system Real-time detection and dynamic optimization of steering performance in unmanned driving mode.
  • the invention solves the problem that the steering wheel is prone to dead zone chattering when the electro-hydraulic intelligent steering system of medium and large commercial vehicles is driving straight, and the steering overshoot is likely to occur when driving in a curve.
  • the problem has broad market application prospects.
  • Figure 1 shows the schematic diagram of the electro-hydraulic intelligent steering system
  • Figure 2 shows the principle diagram of the performance monitor of the electro-hydraulic intelligent steering system.
  • FIG. 1 is a schematic diagram of an electro-hydraulic intelligent steering system.
  • the system includes: steering wheel, recirculating ball steering, electronic control unit (ECU), electric power-assisted module, electro-hydraulic power-assisted module, steering tie rod, wheel.
  • the recirculating ball steering gear includes a steering screw, a steering nut, and a steering gear fan;
  • the electro-hydraulic booster module includes a booster motor, a solenoid valve, a vane pump, and an accumulator.
  • the driver inputs the angle/torque signal to the steering screw through the steering wheel; the electronic control unit (ECU) calculates the electric assist coefficient according to the current working conditions and inputs it to the electric assist module, and the electric assist module outputs the assist signal to the steering screw and combines the torque
  • the signal is output to the steering nut; the electronic control unit (ECU) calculates the hydraulic boost coefficient according to the current working conditions and inputs it to the booster motor and solenoid valve in the electro-hydraulic booster module.
  • the booster motor drives the vane pump to output high-pressure oil signals to the accumulator
  • the valve adjusts the pressure of the flow ⁇ pressure signal output by the accumulator according to the control signal, and outputs the adjusted flow ⁇ pressure signal to the recirculating ball diverter, pushing the steering nut to run, and the steering gear fan converts linear motion into angular motion and outputs To the steering tie rod, so as to drive the wheels to rotate and realize the steering operation.
  • the performance monitor of the electro-hydraulic intelligent steering system of the present invention includes: an information unit, a signal processing unit, a steering performance monitor unit, an execution unit, a steering transmission unit, and a sensing unit ;
  • the information unit includes a driver behavior information module and a vehicle environment information module;
  • the signal processing unit includes a sensor filtering module, an information fusion module and a state estimation module;
  • the steering performance monitor unit includes a steering road feel evaluation subsystem, a steering sensitivity evaluation subsystem, a steering energy consumption evaluation subsystem, and an online optimization solution module;
  • the execution unit includes an electric actuator and a hydraulic actuator
  • the steering transmission unit includes a steering gear module and a steering column module;
  • the sensing unit includes a dynamic parameter sensor and an environment sensing sensor; wherein,
  • the driver transmits the angle/torque signal a to the information unit through the steering wheel.
  • the driver behavior information module and the vehicle environment information module in the information unit respectively output the extracted driver input behavior signal b and surrounding vehicle and environment perception signals c To the signal processing unit; the sensor filter module filters the input signal, the information fusion module fuses the filtered signal, the state estimation module estimates the required unmeasured state variables, and outputs the fused and filtered signal d to the steering Performance monitor unit; the steering road feel evaluation subsystem, steering sensitivity evaluation subsystem, and steering energy consumption evaluation subsystem in the steering performance monitor unit are respectively responsible for steering road feel (the road information felt by the driver) and steering sensitivity (steering wheel).
  • the angle change causes the vehicle yaw rate change) and steering energy consumption to quantify the three steering system evaluation indicators
  • the online optimization solution module quantifies the steering road feel evaluation subsystem, steering sensitivity evaluation subsystem, and steering energy consumption evaluation subsystem.
  • the dynamic optimization calculation is performed to obtain the optimized control signal e and output to the execution unit; the electric actuator and hydraulic actuator in the execution unit output the steering torque f to the steering transmission unit and the steering column according to the optimized control signal e.
  • the module and the steering module sequentially transmit the steering torque f to realize the steering operation; the sensing unit collects the current generated by the steering transmission unit and the rotation angle signal g, and transmits the processed multi-sensor parameter signal h to the information unit , Realize the steering operation of the closed loop system.
  • the multi-sensor parameter signal includes: the yaw rate signal, lateral acceleration signal, pitch angle signal, vehicle speed signal collected by the dynamic parameter sensor, and the distance signal between the main vehicle and the preceding vehicle and the main vehicle collected by the environmental sensor
  • the performance indicators of the steering road feel evaluation subsystem are as follows:
  • ⁇ 0 is the cut-off frequency
  • n e1 is the proportional coefficient between the steering screw angle and the hydraulic actuator motor angle
  • n e2 is the proportional coefficient between the steering screw angle and the electric actuator motor angle
  • l is the center distance of the steering screw
  • a P is the hydraulic cylinder area
  • q is the displacement of the vane pump
  • i g is the transmission ratio
  • ⁇ is the density of the hydraulic oil
  • P is the pitch
  • N is the pump speed
  • C q is the flow coefficient
  • a 1 is the pipe area (assuming the same four pipe cross-section)
  • K S is the stiffness sensor
  • K is the current coefficient
  • n m1 is the motor speed of the hydraulic actuator
  • n m2 for the electric actuator Motor speed
  • n 2 is the reduction ratio of the worm gear of the electric actuator
  • r w is the radius of the pinion gear
  • ⁇ is the speed of the vane pump
  • the performance indicators of the steering sensitivity evaluation subsystem are as follows:
  • m is the mass of the vehicle
  • m s is the sprung mass
  • u is the longitudinal velocity
  • h is the height of the center of mass of the vehicle
  • I x is the moment of inertia around the x-axis of the vehicle
  • I z is the moment of inertia around the z-axis of the vehicle
  • I xz is the moment of inertia around the xz plane of the vehicle
  • L p is the equivalent spring mass coefficient
  • L ⁇ is the equivalent pitch angle coefficient
  • N ⁇ is the wheelbase front wheel rotation angle coefficient
  • N ⁇ is the wheelbase roll angle coefficient
  • N ⁇ is Wheelbase side slip angle coefficient
  • N r is the wheelbase yaw rate coefficient
  • Y r is the suspension yaw rate coefficient
  • Y ⁇ is the suspension front wheel rotation angle coefficient
  • Y ⁇ is the suspension pitch angle coefficient
  • Y ⁇ is the suspension Frame slip angle coefficient
  • a is the distance from the center of
  • the performance indicators of the steering energy consumption evaluation subsystem are as follows:
  • the total energy consumption equation of the electro-hydraulic compound steering system is shown in the following formula.
  • the main power consumption of the system includes four parts: controller power consumption P 1 , motor power consumption P 2 , hydraulic pump power consumption P 3 and rotary valve power consumption P 4 ;
  • E 1 P 1 +P 2 +P 3 +P 4
  • the controller power consumption P 1 is shown in the formula:
  • R A is the armature resistance
  • I A is the armature current
  • U c is the controller voltage
  • R elec is the controller resistance
  • p is the loss during transmission
  • f is the changing frequency of the alternating magnetic field
  • B m is the magnetic density amplitude
  • ⁇ is the Steinmetz coefficient
  • k h , k c and k e are the hysteresis loss coefficient, classical eddy current loss coefficient and eddy current loss coefficient, respectively ;
  • ⁇ is the motor speed
  • Q s is the pump flow
  • P s is the inlet pressure of the power cylinder
  • q is the oil pump displacement
  • ⁇ is the oil density
  • C q is the flow coefficient
  • a 1 and A 2 are the throttling areas of valve ports 1 , 2 respectively
  • Q s is the flow rate of the hydraulic pump
  • Ap is the effective area of the hydraulic cylinder piston
  • x r is The displacement of the steering nut.
  • An electro-hydraulic intelligent steering system performance optimization method of the present invention is based on the above-mentioned monitor and includes the following steps:
  • Step 1 Problem definition, including model definition and algorithm parameter definition
  • Step 2 Initialization: Iterate each particle, assign particle position and velocity, and calculate fitness function until the end of the loop;
  • Step 3 The main loop: The main loop of the algorithm is composed of the basic particle swarm optimization (PSO) module, the adaptive decomposition operator (AD) decomposition module and the Pareto module; first, the particle velocity, position and fitness function values are determined by the basic particles The swarm algorithm module is updated; secondly, the adaptive decomposition operator decomposition module is used to decompose and search the particles updated by the basic particle swarm algorithm module; finally, the loop termination condition is judged, and the Pareto set is derived;
  • PSO basic particle swarm optimization
  • AD adaptive decomposition operator
  • Step 4 If the end condition is not met, the search process returns to step 3, otherwise the Pareto solution set is derived.
  • step 1 specifically includes:
  • Model definition includes: model definition, optimization goals, constraints and design variables;
  • the algorithm parameter definition as compared with MOPSO, not only the maximum number of iterations defined algorithm I te, the number of particles n ori, inertia weight w, the weight decrease rate w damp, individual learning factor c 1, global learning factor c 2, Pareto set
  • the threshold n TPareto also needs to determine the rank of the neighborhood of the decomposition module.
  • step 3 specifically includes:
  • v i,j (t+1) and x i,j (t+1) are the velocity and position of the particle at t+1, respectively, v i,j (t), x i,j (t) Are the speed and position of the particle at time t, c 1 is the individual learning factor, r 1 is the individual learning factor weight, c 2 is the global learning factor, r 2 is the global learning factor weight, p i, j are the current search process , P g,j is the current global optimal particle;
  • AD Adaptive decomposition operator
  • the decomposition search strategy is divided into two parts. First, the neighbor of the current sub-problem is determined by the distance matrix to operate the decomposition search; then, an adaptive search is performed in each search direction determined by the weight matrix.
  • the adaptive search process mainly includes: Establish a multi-objective optimization model with d 1 (p) and d 2 (p) as optimization goals and algorithm robustness criteria as constraints. By adjusting the design variable p, when the optimization model meets the robustness constraints, d 1 ( p) and d 2 (p) tend to be the smallest; the optimization model is shown in the following formula:
  • d 1 (p) is the particle convergence distance
  • d 2 (p) is the particle diversity distance
  • p is the design variable of the decomposition search module
  • w, p) is the search method
  • x* is the ideal reference point
  • x k is the point obtained by the actual search
  • w is the weight coefficient of the decomposition search module
  • x and y be the two solution points obtained after one iteration, let x.object(j) denote the fitness function value of object j corresponding to particle x; for the multi-objective optimization problem with the smallest objective value, if all x.object (j) is less than or equal to y.object(j), and at least one x.object(j) is less than y.object(j), then x belongs to the Pareto solution set.

Landscapes

  • Engineering & Computer Science (AREA)
  • Chemical & Material Sciences (AREA)
  • Combustion & Propulsion (AREA)
  • Transportation (AREA)
  • Mechanical Engineering (AREA)
  • Steering Control In Accordance With Driving Conditions (AREA)

Abstract

一种电液智能转向系统性能监测器及性能优化方法，通过信息单元对采集到的车辆动力学参数信号、周边车辆及道路环境感知信号以及驾驶员输入行为信号进行处理，信号处理单元将分析后的结果传递至转向性能监测器单元，所述转向系统性能监测单元包括转向路感、转向灵敏度以及转向能耗三个评价子系统，可根据评价方程对当前控制信号进行在线优化解算，从而获取最佳转向行为，再将优化后的控制信号输入至执行单元，从而实现了系统在人机共驾下转向性能的实时检测及动态优化，解决了中大型商用车电液智能转向系统直线行驶时转向盘易发生死区抖震、弯道行驶时易出现转向超调等问题，具有广阔的市场应用前景。

Description

一种电液智能转向系统性能监测器及性能优化方法 技术领域
本发明属于汽车助力转向系统控制技术领域，具体指代一种电液智能转向系统性能监测器及性能优化方法。
背景技术
智能化是未来汽车发展的趋势，智能转向系统作为智能网联车辆控制中重要的组成部分，其性能的实时监测技术和动态优化问题逐渐受到了广泛的关注。电液智能转向系统是一种用于大中型商用车的新型转向系统，其特点在于可通过协调双执行机构的输出比例实现转向模式的切换，即在低车速时采用大转矩输出的复合转向模式，高车速时采用路感良好且更为节能的电动转向模式，是一种理想的转向系统设计形式。
然而，电液智能转向系统中不同的参数值对系统的性能有很大的影响，因此需要通过对转向系统动力学模型进行分析得到包含其关键结构参数的性能指标量化公式，从而能够在设计阶段对转向系统的性能进行提升。对于新型电液复合转向系统的性能优化设计包含两方面的问题：一是如何根据待求解模型的设计特点和动力学特性，提出待优化系统的评价指标和约束条件，从而进行多目标优化模型的构建；二是根据所构建问题特点，设计相应数值优化求解算法，对所提出问题进行计算。
第一，在对于现有转向系统性能指标提出和量化的研究中，赵万忠等在论文《Integrated optimization of active steering and semi-active suspension based on an improved memetic algorithm》和《Primary studies on integration optimization of differential steering of electric vehicle with motorized wheels based on quality engineering》中提出通过力矩传递函数与角位移传递函数对电动助力转向系统以及差速助力转向系统性能进行定量分析，根据转向系统的设计需求提出了转向路感、转向灵敏度等系统性能评价指标作为优化目标和约束条件，从而构建包含系统关键结构参数的优化模型，将汽车转向系统性能优化问题转换为多目标参数优化问题。王晓晶等在论文《基于物理规划与遗传算法转向系统优化设计》中提出针对转向系统优化问题，以满载低速转向时转角偏差、空载高速转向时转角偏差、满载与空载的车轮跳动时轮胎滑移量及车轮摆角摆动误差为目标建立多目标优化模型，并通过物理规划法构造所建立数学模型各优化目标的偏好等级。解后循等基于电动液压助力转向系统液压阻尼系数及阻力系数对电动液压助力转向系统阻尼及助力系数进行量化，并提出以方向盘路感最优为目标函数的优化模型。然而，上述研究均只针对传统车辆操纵稳定性需求为目标进行性能指标量化，未将目前电动车辆节能需求以及电液智能转向系统的性能需求考虑在内，并提出相应的评价指标。
第二，在优化算法方面，对于单目标优化问题的求解来说，理论上必然可获得唯一的最优解，然而对于多目标优化问题来说，一个性能指标提高可能导致另一个性能指标的降低。在多目标问题的求解方法中，加权和法是一种常用的多目标优化问题求解方法，其原理为通 过权重因子、比例因子将多个目标整合为一个目标，即将多目标优化问题转换为一个单目标优化问题，再通过单目标优化的方法进行计算，该方法在工程中简单实用，可直接得到一个最优解，但若其中一个目标过于敏感，可能导致其他目标的优化结果较差。相比传统多目标优化数值求解算法，多目标进化算法的特点在于以“种群”作为进化载体，从而可实现并行计算。算法基于“适应度”来反映新解的优劣程度并通过交叉、变异等更新策略，从而保证新种群的多样性，防止求解过程陷入局部最优。目前常用的多目标进化算法主要基于支配策略进行设计，其特点在于引入了占优和支配的概念，但其存在求解过程后期非支配解过多从而影响求解效率的问题。不同于上述基于支配策略的多目标优化算法，基于分解策略进行设计的MOEA/D算法的求解原理为通过“聚合函数”将一个多目标优化(Multi-objective optimization，MOP)问题分解为多个单目标优化(Single-objective optimization，SOP)问题，再通过进化算法在一次迭代中同时求解基于分解策略得到的所有SOP子问题，并通过在目标空间中分布特性良好的权向量来保证整个求解过程的种群多样性，即通过“邻域”来使得种群中的个体都以“小团体”的方式相互协同进化。相比其他MOEA算法，该策略既降低了算法的时间复杂度，又提高了解的精度，但现有的固定搜索策略不能实现自适应调整，从而在求解实际工程优化问题时存在效率问题。
为了提高电液复合转向系统的性能，实现降低转向能耗、提高转向操纵稳定性的优化设计目标，需要综合考虑系统的动力学特性、功能设计需求以及系统所匹配的整车特性，建立可定量描述系统性能的量化公式，从而将系统的优化设计问题转化为多目标参数优化问题。但由于工程问题优化模型一般为非线性、离散的，采用基本算法难以逼近优化模型的帕累托前沿。因此，需要针对问题特点对算法进行改进，并设计算例对算法性能进行测试，从而保证在求解由系统设计需求转换得到的优化模型时的高效性。此外，在算法改进方面，基于分解策略的MOEA/D多目标优化算法架构是一种计算复杂度低、求解精度高的多目标问题求解方法，以PSO算子为代表的群智能算法具有鲁棒性强、对环境变化具有自适应性和快速反应性等特点，适合用于工程优化问题的求解计算，是一种理想转向系统性能动态优化方法，具有广阔的应用前景。
发明内容
针对于上述现有技术的不足，本发明的目的在于提供一种电液智能转向系统性能监测器及性能优化方法，以解决现有技术中不能对当前转向系统性能进行实时在线监测以及动态优化解算等问题。
为达到上述目的，本发明采用的技术方案如下：
本发明的一种电液智能转向系统性能监测器，包括：信息单元、信号处理单元、转向性能监测器单元、执行单元、转向传动单元和感知单元；
所述信息单元包括驾驶员行为信息模块和车辆环境信息模块；
所述信号处理单元包括传感器滤波模块、信息融合模块和状态估计模块；
所述转向性能监测器单元包括转向路感评价子系统、转向灵敏度评价子系统、转向能耗评价子系统和在线优化解算模块；
所述执行单元包括电动执行机构和液压执行机构；
所述转向传动单元包括转向器模块和转向管柱模块；
所述感知单元包括动力学参数传感器和环境感知传感器；其中，
驾驶员通过转向盘将转角/转矩信号a传递给信息单元，信息单元中驾驶员行为信息模块和车辆环境信息模块分别将提取得到的驾驶员输入行为信号b和周边车辆及环境感知信号c输出给信号处理单元；传感器滤波模块对输入信号进行滤波处理，信息融合模块对滤波后的信号进行融合，状态估计模块对所需不可测状态变量进行估计计算，将融合、滤波后信号d输出给转向性能监测器单元；转向性能监测器单元中的转向路感评价子系统、转向灵敏度评价子系统、转向能耗评价子系统分别对转向路感(驾驶员感受到的路面信息)、转向灵敏度(方向盘角度变化引起车辆横摆角速度变化)和转向能耗三项转向系统评价指标进行量化，在线优化解算模块对转向路感评价子系统、转向灵敏度评价子系统、转向能耗评价子系统得到的量化结果进行动态优化计算，得到优化后的控制信号e并输出给执行单元；执行单元中的电动执行机构和液压执行机构根据上述优化后的控制信号e输出转向力矩f至转向传动单元，转向管柱模块、转向器模块依次对转向力矩f进行传递，实现转向操作；所述感知单元对转向传动单元产生的电流、转角信号g进行采集，并将经过处理后的多传感器参数信号h传递至信息单元，实现闭环系统的转向操作。
进一步地，所述多传感器参数信号包括：动力学参数传感器采集到的横摆角速度信号、侧向加速度信号、俯仰角信号、车速信号，及环境感知传感器采集到的主车与前车距离信号、主车与前车相对速度信号、左侧车道车辆与主车距离信号、左侧车道车辆与主车相对速度信号、右侧车道车辆与主车距离信号、右侧车道车辆与主车相对速度信号。
进一步地，所述转向路感评价子系统性能指标如下：
Figure PCTCN2019116038-appb-000001
Figure PCTCN2019116038-appb-000002
式中，ω 0为截止频率，n e1为转向螺杆角与液压执行机构马达角的比例系数，n e2为转向 螺杆角与电动执行机构马达角的比例系数，l为转向螺杆的中心距离，A P为液压缸面积，q为叶片泵的排量，i g为传动比，ρ为液压油的密度，P为螺距，N为泵的转速，C q为流量系数，A 1为管道面积(假设四个管道横截面相同)，K s为传感器刚度，K为电流系数，K a为电动执行机构的电机电枢扭矩电流系数，n m1为液压执行机构的电机速度，n m2为电动执行机构的电机速度，n 2为电动执行机构的蜗轮减速比，r w为小齿轮半径，ω为叶片泵转速，
Figure PCTCN2019116038-appb-000003
为转向螺杆角加速度，j为虚数单位，J m2为电动执行机构的电机转动惯量，J m1为液压执行机构的电机转动惯量，B lg为转向螺杆的阻尼，B lm为转向螺母阻尼的阻尼，B cs为转向齿扇的阻尼，J cs为转向齿扇的转动惯量。
进一步地，所述转向灵敏度评价子系统性能指标如下：
Figure PCTCN2019116038-appb-000004
Figure PCTCN2019116038-appb-000005
Figure PCTCN2019116038-appb-000006
Figure PCTCN2019116038-appb-000007
Figure PCTCN2019116038-appb-000008
Figure PCTCN2019116038-appb-000009
式中，m为车辆质量，m s为簧载质量，u为纵向速度，h为车辆的质心高度，I x为绕车辆x轴的惯性矩，I z为绕车辆z轴的惯性矩，I xz为围绕车辆x-z平面的惯性矩，L p为等效弹簧质量系数，L φ为等效俯仰角系数，N δ为轴距前轮转角系数，N φ为轴距侧倾角系数，N β为轴距侧偏角系数，N r为轴距横摆角速度系数，Y r为悬架横摆角速度系数，Y δ为悬架前轮转角系数，Y φ为悬架俯仰角系数，Y β为悬架侧偏角系数，a为从车辆质心到前轴的距离，k 1为前轮侧偏刚度，E 1为前侧倾转向系数。
进一步地，所述转向能耗评价子系统性能指标如下：
电液复合转向系统的总能耗方程如下式所示，系统的主要功耗包括控制器功耗P 1、电机功耗P 2、液压泵功耗P 3和旋转阀功耗P 4四个部分；
E 1＝P 1+P 2+P 3+P 4
所述控制器功耗P 1如式所示：
Figure PCTCN2019116038-appb-000010
式中，R A为电枢电阻，I A为电枢电流，U c为控制器电压，R elec为控制器电阻，p为传 输过程中的损耗；
所述电机功率损耗原理为，考虑到铁损是电机的主要能耗，根据分离电机铁损的原理，建立了电机能耗的计算模型，如下式所示：
P 2＝k hfB m α+k cfB m+k ef 1.5B m 1.5
式中，f为交变磁场的变化频率，B m为磁密度幅值，α为斯坦梅兹系数，k h、k c和k e分别为磁滞损耗系数、经典涡流损耗系数和涡流损耗系数；
所述泵的功率损耗原理如下式所示：
Figure PCTCN2019116038-appb-000011
式中，ω为马达转速，Q s为泵流量，P s为动力缸的入口压力，q为油泵排量；
所述转阀功率损耗原理如下式所示：
Figure PCTCN2019116038-appb-000012
式中，ρ为油密度，C q为流量系数，A 1、A 2分别为1、2阀口节流面积，Q s为液压泵的流量，A p为液压缸活塞有效面积，x r为转向螺母的位移。
本发明的一种电液智能转向系统性能优化方法，包括步骤如下：
步骤1：问题定义，包括模型定义和算法参数定义；
步骤2：初始化：迭代每个粒子，分配粒子位置和速度，并计算适应度函数直到循环结束；
步骤3：主循环：算法主回路由基本粒子群算法(PSO)模块、自适应分解算子(AD)分解模块和Pareto模块三部分组成；首先，粒子速度、位置和适应度函数值由基本粒子群算法模块更新；其次，自适应分解算子分解模块用于对基本粒子群算法模块更新的粒子进行分解搜索；最后，判断环路终止条件，导出帕累托集；
步骤4：如果不满足结束条件，则搜索过程返回步骤3，否则导出Pareto解集。
进一步地，所述步骤1具体包括：
11)模型定义包括：模型的定义、优化目标、约束条件和设计变量；
12)算法参数定义，与MOPSO相比，不仅定义算法I te的最大迭代次数、粒子数n ori、惯性权重w、权重下降率w damp、个体学习因子c 1、全局学习因子c 2、Pareto集的阈值n TPareto，还需确定分解模块邻域的位阶。
进一步地，所述步骤3具体包括：
31)基本粒子群算法模块
更新粒子速度、位置和适应度函数值，如下：
Figure PCTCN2019116038-appb-000013
式中，v i,j(t+1)、x i,j(t+1)分别为粒子在t+1时刻的速度和位置，v i,j(t)、x i,j(t)分别为粒子在t时刻的速度和位置，c 1为个体学习因子，r 1为个体学习因子权重，c 2为全局学习因子，r 2为全局学习因子权重，p i,j为当前搜索过程中的粒子，p g,j为当前全局最优粒子；
32)自适应分解算子(AD)分解模块搜索策略
分解搜索策略分为两部分，首先，通过距离矩阵确定当前子问题的邻居来操作分解搜索；然后，在权重矩阵确定的每个搜索方向上执行自适应搜索，所述自适应搜索过程主要包括，建立以d 1(p)和d 2(p)为优化目标、以算法鲁棒性准则为约束的多目标优化模型，通过调整设计变量p，当优化模型满足鲁棒性约束时，d 1(p)和d 2(p)趋于最小；优化模型如下式所示：
Figure PCTCN2019116038-appb-000014
式中，d 1(p)为粒子收敛距离，d 2(p)为粒子分集距离，p为分解搜索模块设计变量，g wd(x*|w,p)、g wd(x k|w,p)为搜索方式，x*为理想参考点，x k为实际搜索得到的点，w为分解搜索模块权重系数；
3)Pareto模块策略
设x和y是一次迭代后得到的两个解点，令x.object(j)表示对象j对应粒子x的适应度函数值；对于具有最小目标值的多目标优化问题，如果所有x.object(j)小于或等于y.object(j)，并且至少一个x.object(j)小于y.object(j)，则x属于Pareto解集。
本发明的有益效果：
本发明通过传感器滤波模块、信息融合模块和状态估计模块对采集到的车辆信号、环境信号以及驾驶员信号进行处理分析后传递至性能监测器，所述性能监测器包括转向路感、转向灵敏度以及转向能耗三个评价子系统，可根据所述评价方程对当前系统性能进行在线优化解算，从而获取最佳转向行为决策，再将优化后的控制信号输入至下层执行单元，从而实现了系统在无人驾驶模式下转向性能的实时检测及动态优化，本发明解决了中大型商用车电液智能转向系统直线行驶时转向盘易发生死区抖震、弯道行驶时易出现转向超调等问题，具有广阔的市场应用前景。
附图说明
图1绘示电液智能转向系统的原理图；
图2绘示电液智能转向系统性能监测器的原理图。
具体实施方式
为了便于本领域技术人员的理解，下面结合实施例与附图对本发明作进一步的说明，实施方式提及的内容并非对本发明的限定。
参照图1所示，其为电液智能转向系统的原理图，该系统，包括：转向盘、循环球转向器、电子控制单元(ECU)、电动助力模块、电液助力模块、转向横拉杆、车轮。其中，所述循环球转向器包括转向螺杆、转向螺母、转向齿扇；所述电液助力模块包括助力电机、电磁阀、叶片泵、蓄能器。
驾驶员通过转向盘输入转角/转矩信号到转向螺杆；电子控制单元(ECU)根据当前工况计算得到电助力系数输入到电动助力模块，电动助力模块输出助力信号到转向螺杆，并将合力矩信号输出到转向螺母；电子控制单元(ECU)根据当前工况计算得到液助力系数输入到电液助力模块中的助力电机和电磁阀，助力电机驱动叶片泵输出高压油信号到蓄能器，电磁阀根据控制信号对蓄能器输出的流量\压力信号进行调压，将调节后的流量\压力信号输出到循环球转向器，推动转向螺母运行，转向齿扇将直线运动转换成转角运动并输出到转向横拉杆，从而带动车轮转动，实现转向操作。
参照图2所示，本发明的电液智能转向系统性能监测器(即上述的电子控制单元)，包括：信息单元、信号处理单元、转向性能监测器单元、执行单元、转向传动单元和感知单元；
所述信息单元包括驾驶员行为信息模块和车辆环境信息模块；
所述信号处理单元包括传感器滤波模块、信息融合模块和状态估计模块；
所述转向性能监测器单元包括转向路感评价子系统、转向灵敏度评价子系统、转向能耗评价子系统和在线优化解算模块；
所述执行单元包括电动执行机构和液压执行机构；
所述转向传动单元包括转向器模块和转向管柱模块；
所述感知单元包括动力学参数传感器和环境感知传感器；其中，
驾驶员通过转向盘将转角/转矩信号a传递给信息单元，信息单元中驾驶员行为信息模块和车辆环境信息模块分别将提取得到的驾驶员输入行为信号b和周边车辆及环境感知信号c输出给信号处理单元；传感器滤波模块对输入信号进行滤波处理，信息融合模块对滤波后的信号进行融合，状态估计模块对所需不可测状态变量进行估计计算，将融合、滤波后信号d输出给转向性能监测器单元；转向性能监测器单元中的转向路感评价子系统、转向灵敏度评价子系统、转向能耗评价子系统分别对转向路感(驾驶员感受到的路面信息)、转向灵敏度(方向盘角度变化引起车辆横摆角速度变化)和转向能耗三项转向系统评价指标进行量化，在线优化解算模块对转向路感评价子系统、转向灵敏度评价子系统、转向能耗评价子系统得到的量化结果进行动态优化计算，得到优化后的控制信号e并输出给执行单元；执行单元中的电动执行机构和液压执行机构根据上述优化后的控制信号e输出转向力矩f至转向传动单元，转向管柱模块、转向器模块依次对转向力矩f进行传递，实现转向操作；所述感知单元对转 向传动单元产生的电流、转角信号g进行采集，并将经过处理后的多传感器参数信号h传递至信息单元，实现闭环系统的转向操作。
其中，所述多传感器参数信号包括：动力学参数传感器采集到的横摆角速度信号、侧向加速度信号、俯仰角信号、车速信号，及环境感知传感器采集到的主车与前车距离信号、主车与前车相对速度信号、左侧车道车辆与主车距离信号、左侧车道车辆与主车相对速度信号、右侧车道车辆与主车距离信号、右侧车道车辆与主车相对速度信号。
其中，所述转向路感评价子系统性能指标如下：
Figure PCTCN2019116038-appb-000015
Figure PCTCN2019116038-appb-000016
式中，ω 0为截止频率，n e1为转向螺杆角与液压执行机构马达角的比例系数，n e2为转向螺杆角与电动执行机构马达角的比例系数，l为转向螺杆的中心距离，A P为液压缸面积，q为叶片泵的排量，i g为传动比，ρ为液压油的密度，P为螺距，N为泵的转速，C q为流量系数，A 1为管道面积(假设四个管道横截面相同)，K s为传感器刚度，K为电流系数，K a为电动执行机构的电机电枢扭矩电流系数，n m1为液压执行机构的电机速度，n m2为电动执行机构的电机速度，n 2为电动执行机构的蜗轮减速比，r w为小齿轮半径，ω为叶片泵转速，
Figure PCTCN2019116038-appb-000017
为转向螺杆角加速度，j为虚数单位，J m2为电动执行机构的电机转动惯量，J m1为液压执行机构的电机转动惯量，B lg为转向螺杆的阻尼，B lm为转向螺母阻尼的阻尼，B cs为转向齿扇的阻尼，J cs为转向齿扇的转动惯量。
其中，所述转向灵敏度评价子系统性能指标如下：
Figure PCTCN2019116038-appb-000018
Figure PCTCN2019116038-appb-000019
Figure PCTCN2019116038-appb-000020
Figure PCTCN2019116038-appb-000021
Figure PCTCN2019116038-appb-000022
Figure PCTCN2019116038-appb-000023
式中，m为车辆质量，m s为簧载质量，u为纵向速度，h为车辆的质心高度，I x是绕车辆x轴的惯性矩，I z为绕车辆z轴的惯性矩，I xz为围绕车辆x-z平面的惯性矩，L p为等效弹簧质量系数，L φ为等效俯仰角系数，N δ为轴距前轮转角系数，N φ为轴距侧倾角系数，N β为 轴距侧偏角系数，N r为轴距横摆角速度系数，Y r为悬架横摆角速度系数，Y δ为悬架前轮转角系数，Y φ为悬架俯仰角系数，Y β为悬架侧偏角系数，a为从车辆质心到前轴的距离，k 1为前轮侧偏刚度，E 1为前侧倾转向系数。
其中，所述转向能耗评价子系统性能指标如下：
电液复合转向系统的总能耗方程如下式所示，系统的主要功耗包括控制器功耗P 1、电机功耗P 2、液压泵功耗P 3和旋转阀功耗P 4四个部分；
E 1＝P 1+P 2+P 3+P 4
所述控制器功耗P 1如式所示：
Figure PCTCN2019116038-appb-000024
式中，R A为电枢电阻，I A为电枢电流，U c为控制器电压，R elec为控制器电阻，p为传输过程中的损耗；
所述电机功率损耗原理为，考虑到铁损是电机的主要能耗，根据分离电机铁损的原理，建立了电机能耗的计算模型，如下式所示：
P 2＝k hfB m α+k cfB m+k ef 1.5B m 1.5
式中，f为交变磁场的变化频率，B m为磁密度幅值，α为斯坦梅兹系数，k h、k c和k e分别为磁滞损耗系数、经典涡流损耗系数和涡流损耗系数；
所述泵的功率损耗原理如下式所示：
Figure PCTCN2019116038-appb-000025
式中，ω为马达转速，Q s为泵流量，P s为动力缸的入口压力，q是油泵排量；
所述转阀功率损耗原理如下式所示：
Figure PCTCN2019116038-appb-000026
式中，ρ为油密度，C q为流量系数，A 1、A 2分别为1、2阀口节流面积，Q s为液压泵的流量，A p为液压缸活塞有效面积，x r为转向螺母的位移。
本发明的一种电液智能转向系统性能优化方法，基于上述监测器，包括步骤如下：
步骤1：问题定义，包括模型定义和算法参数定义；
步骤2：初始化：迭代每个粒子，分配粒子位置和速度，并计算适应度函数直到循环结 束；
步骤3：主循环：算法主回路由基本粒子群算法(PSO)模块、自适应分解算子(AD)分解模块和Pareto模块三部分组成；首先，粒子速度、位置和适应度函数值由基本粒子群算法模块更新；其次，自适应分解算子分解模块用于对基本粒子群算法模块更新的粒子进行分解搜索；最后，判断环路终止条件，导出帕累托集；
步骤4：如果不满足结束条件，则搜索过程返回步骤3，否则导出Pareto解集。
其中，所述步骤1具体包括：
11)模型定义包括：模型的定义、优化目标、约束条件和设计变量；
12)算法参数定义，与MOPSO相比，不仅定义算法I te的最大迭代次数、粒子数n ori、惯性权重w、权重下降率w damp、个体学习因子c 1、全局学习因子c 2、Pareto集的阈值n TPareto，还需确定分解模块邻域的位阶。
其中，所述步骤3具体包括：
31)基本粒子群算法模块
更新粒子速度、位置和适应度函数值，如下：
Figure PCTCN2019116038-appb-000027
式中，v i,j(t+1)、x i,j(t+1)分别为粒子在t+1时刻的速度和位置，v i,j(t)、x i,j(t)分别为粒子在t时刻的速度和位置，c 1为个体学习因子，r 1为个体学习因子权重，c 2为全局学习因子，r 2为全局学习因子权重，p i,j为当前搜索过程中的粒子，p g,j为当前全局最优粒子；
32)自适应分解算子(AD)分解模块搜索策略
分解搜索策略分为两部分，首先，通过距离矩阵确定当前子问题的邻居来操作分解搜索；然后，在权重矩阵确定的每个搜索方向上执行自适应搜索，所述自适应搜索过程主要包括，建立以d 1(p)和d 2(p)为优化目标、以算法鲁棒性准则为约束的多目标优化模型，通过调整设计变量p，当优化模型满足鲁棒性约束时，d 1(p)和d 2(p)趋于最小；优化模型如下式所示：
Figure PCTCN2019116038-appb-000028
式中，d 1(p)为粒子收敛距离，d 2(p)为粒子分集距离，p为分解搜索模块设计变量，g wd(x*|w,p)、g wd(x k|w,p)为搜索方式，x*为理想参考点，x k为实际搜索得到的点，w为分解搜索模块权重系数；
3)Pareto模块策略
设x和y是一次迭代后得到的两个解点，令x.object(j)表示对象j对应粒子x的适应度函数值；对于具有最小目标值的多目标优化问题，如果所有x.object(j)小于或等于y.object(j)， 并且至少一个x.object(j)小于y.object(j)，则x属于Pareto解集。
本发明具体应用途径很多，以上所述仅是本发明的优选实施方式，应当指出，对于本技术领域的普通技术人员来说，在不脱离本发明原理的前提下，还可以作出若干改进，这些改进也应视为本发明的保护范围。

Claims (8)

  1. 一种电液智能转向系统性能监测器，其特征在于,包括：信息单元、信号处理单元、转向性能监测器单元、执行单元、转向传动单元和感知单元；
    所述信息单元包括驾驶员行为信息模块和车辆环境信息模块；
    所述信号处理单元包括传感器滤波模块、信息融合模块和状态估计模块；
    所述转向性能监测器单元包括转向路感评价子系统、转向灵敏度评价子系统、转向能耗评价子系统和在线优化解算模块；
    所述执行单元包括电动执行机构和液压执行机构；
    所述转向传动单元包括转向器模块和转向管柱模块；
    所述感知单元包括动力学参数传感器和环境感知传感器；其中，
    驾驶员通过转向盘将转角/转矩信号a传递给信息单元，信息单元中驾驶员行为信息模块和车辆环境信息模块分别将提取得到的驾驶员输入行为信号b和周边车辆及环境感知信号c输出给信号处理单元；传感器滤波模块对输入信号进行滤波处理，信息融合模块对滤波后的信号进行融合，状态估计模块对所需不可测状态变量进行估计计算，将融合、滤波后信号d输出给转向性能监测器单元；转向性能监测器单元中的转向路感评价子系统、转向灵敏度评价子系统、转向能耗评价子系统分别对转向路感、转向灵敏度和转向能耗三项转向系统评价指标进行量化，在线优化解算模块对转向路感评价子系统、转向灵敏度评价子系统、转向能耗评价子系统得到的量化结果进行动态优化计算，得到优化后的控制信号e并输出给执行单元；执行单元中的电动执行机构和液压执行机构根据上述优化后的控制信号e输出转向力矩f至转向传动单元，转向管柱模块、转向器模块依次对转向力矩f进行传递，实现转向操作；所述感知单元对转向传动单元产生的电流、转角信号g进行采集，并将经过处理后的多传感器参数信号h传递至信息单元，实现闭环系统的转向操作。
  2. 根据权利要求1所述的电液智能转向系统性能监测器，其特征在于，所述多传感器参数信号包括：动力学参数传感器采集到的横摆角速度信号、侧向加速度信号、俯仰角信号、车速信号，及环境感知传感器采集到的主车与前车距离信号、主车与前车相对速度信号、左侧车道车辆与主车距离信号、左侧车道车辆与主车相对速度信号、右侧车道车辆与主车距离信号、右侧车道车辆与主车相对速度信号。
  3. 根据权利要求1所述的电液智能转向系统性能监测器，其特征在于，所述转向路感评价子系统性能指标如下：
    Figure PCTCN2019116038-appb-100001
    Figure PCTCN2019116038-appb-100002
    式中，ω 0为截止频率，n e1为转向螺杆角与液压执行机构马达角的比例系数，n e2为转向螺杆角与电动执行机构马达角的比例系数，l为转向螺杆的中心距离，A P为液压缸面积，q为叶片泵的排量，i g为传动比，ρ为液压油的密度，P为螺距，N为泵的转速，C q为流量系数，A 1为管道面积，K s为传感器刚度，K为电流系数，K a为电动执行机构的电机电枢扭矩电流系数，n m1为液压执行机构的电机速度，n m2为电动执行机构的电机速度，n 2为电动执行机构的蜗轮减速比，r w为小齿轮半径，ω为叶片泵转速，
    Figure PCTCN2019116038-appb-100003
    为转向螺杆角加速度，j为虚数单位，J m2为电动执行机构的电机转动惯量，J m1为液压执行机构的电机转动惯量，B lg为转向螺杆的阻尼，B lm为转向螺母阻尼的阻尼，B cs为转向齿扇的阻尼，J cs为转向齿扇的转动惯量。
  4. 根据权利要求1所述的电液智能转向系统性能监测器，其特征在于，所述转向灵敏度评价子系统性能指标如下：
    Figure PCTCN2019116038-appb-100004
    Figure PCTCN2019116038-appb-100005
    Figure PCTCN2019116038-appb-100006
    Figure PCTCN2019116038-appb-100007
    Figure PCTCN2019116038-appb-100008
    Figure PCTCN2019116038-appb-100009
    式中，m为车辆质量，m s为簧载质量，u为纵向速度，h为车辆的质心高度，I x为绕车辆x轴的惯性矩，I z为绕车辆z轴的惯性矩，I xz为围绕车辆x-z平面的惯性矩，L p为等效弹簧质量系数，L φ为等效俯仰角系数，N δ为轴距前轮转角系数，N φ为轴距侧倾角系数，N β为轴距侧偏角系数，N r为轴距横摆角速度系数，Y r为悬架横摆角速度系数，Y δ为悬架前轮转角系数，Y φ为悬架俯仰角系数，Y β为悬架侧偏角系数，a为从车辆质心到前轴的距离，k 1为前轮侧偏刚度，E 1为前侧倾转向系数。
  5. 根据权利要求1所述的电液智能转向系统性能监测器，其特征在于，所述转向能耗评价子系统性能指标如下：
    电液复合转向系统的总能耗方程如下式所示，系统的功耗包括控制器功耗P 1、电机功耗P 2、液压泵功耗P 3和旋转阀功耗P 4四个部分；
    E 1＝P 1+P 2+P 3+P 4
    所述控制器功耗P 1如式所示：
    Figure PCTCN2019116038-appb-100010
    式中，R A为电枢电阻，I A为电枢电流，U c为控制器电压，R elec为控制器电阻，p为传输过程中的损耗；
    所述电机功率损耗原理为，考虑到铁损是电机的主要能耗，根据分离电机铁损的原理，建立了电机能耗的计算模型，如下式所示：
    P 2＝k hfB m α+k cfB m+k ef 1.5B m 1.5
    式中，f为交变磁场的变化频率，B m为磁密度幅值，α为斯坦梅兹系数，k h、k c和k e分别为磁滞损耗系数、经典涡流损耗系数和涡流损耗系数；
    所述泵的功率损耗原理如下式所示：
    Figure PCTCN2019116038-appb-100011
    式中，ω为马达转速，Q s为泵流量，P s为动力缸的入口压力，q为油泵排量；
    所述转阀功率损耗原理如下式所示：
    Figure PCTCN2019116038-appb-100012
    式中，ρ为油密度，C q为流量系数，A 1、A 2分别为1、2阀口节流面积，Q s为液压泵的流量，A p为液压缸活塞有效面积，x r为转向螺母的位移。
  6. 一种电液智能转向系统性能优化方法，其特征在于，包括步骤如下：
    步骤1：问题定义，包括模型定义和算法参数定义；
    步骤2：初始化：迭代每个粒子，分配粒子位置和速度，并计算适应度函数直到循环结束；
    步骤3：主循环：算法主回路由基本粒子群算法模块、自适应分解算子分解模块和Pareto模块三部分组成；首先，粒子速度、位置和适应度函数值由基本粒子群算法模块更新；其次，自适应分解算子分解模块用于对基本粒子群算法模块更新的粒子进行分解搜索；最后，判断环路终止条件，导出帕累托集；
    步骤4：如果不满足结束条件，则搜索过程返回步骤3，否则导出Pareto解集。
  7. 根据权利要求6所述的电液智能转向系统性能优化方法，其特征在于，所述步骤1具体包括：
    11)模型定义包括：模型的定义、优化目标、约束条件和设计变量；
    12)算法参数定义，不仅定义算法I te的最大迭代次数、粒子数n ori、惯性权重w、权重下降率w damp、个体学习因子c 1、全局学习因子c 2、Pareto集的阈值n TPareto，还需确定分解模块邻域的位阶。
  8. 根据权利要求6所述的电液智能转向系统性能优化方法，其特征在于，所述步骤3具体包括：
    31)基本粒子群算法模块
    更新粒子速度、位置和适应度函数值，如下：
    Figure PCTCN2019116038-appb-100013
    式中，v i,j(t+1)、x i,j(t+1)分别为粒子在t+1时刻的速度和位置，v i,j(t)、x i,j(t)分别为粒子在t时刻的速度和位置，c 1为个体学习因子，r 1为个体学习因子权重，c 2为全局学习因子，r 2为全局学习因子权重，p i,j为当前搜索过程中的粒子，p g,j为当前全局最优粒子；
    32)自适应分解算子分解模块搜索策略
    分解搜索策略分为两部分，首先，通过距离矩阵确定当前子问题的邻居来操作分解搜索；然后，在权重矩阵确定的每个搜索方向上执行自适应搜索，所述自适应搜索过程主要包括，建立以d 1(p)和d 2(p)为优化目标、以算法鲁棒性准则为约束的多目标优化模型，通过调整设计变量p，当优化模型满足鲁棒性约束时，d 1(p)和d 2(p)趋于最小；优化模型如下式所示：
    Figure PCTCN2019116038-appb-100014
    式中，d 1(p)为粒子收敛距离，d 2(p)为粒子分集距离，p为分解搜索模块设计变量，g wd(x*|w,p)、g wd(x k|w,p)为搜索方式，x*为理想参考点，x k为实际搜索得到的点，w为分解搜索模块权重系数；
    3)Pareto模块策略
    设x和y是一次迭代后得到的两个解点，令x.object(j)表示对象j对应粒子x的适应度函数值；对于具有最小目标值的多目标优化问题，如果所有x.object(j)小于或等于y.object(j)，并且至少一个x.object(j)小于y.object(j)，则x属于Pareto解集。
PCT/CN2019/116038 2019-04-23 2019-11-06 一种电液智能转向系统性能监测器及性能优化方法 WO2020215659A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910327314.5A CN110386191B (zh) 2019-04-23 2019-04-23 一种电液智能转向系统性能监测器及性能优化方法
CN201910327314.5 2019-04-23

Publications (1)

Publication Number Publication Date
WO2020215659A1 true WO2020215659A1 (zh) 2020-10-29

Family

ID=68284977

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/116038 WO2020215659A1 (zh) 2019-04-23 2019-11-06 一种电液智能转向系统性能监测器及性能优化方法

Country Status (2)

Country Link
CN (1) CN110386191B (zh)
WO (1) WO2020215659A1 (zh)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110386191B (zh) * 2019-04-23 2020-07-24 南京航空航天大学 一种电液智能转向系统性能监测器及性能优化方法
CN113200056B (zh) * 2021-06-22 2022-04-08 吉林大学 一种非完全信息非合作博弈人机共驾控制方法
CN113602348B (zh) * 2021-08-27 2022-10-28 奇瑞新能源汽车股份有限公司 转向助力曲线调校方法、装置、车辆及存储介质
CN117708997B (zh) * 2024-02-05 2024-05-03 中汽研汽车检验中心（天津）有限公司 基于遗传算法的车辆悬架设计参数确定方法、设备和介质

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103019822A (zh) * 2012-12-07 2013-04-03 北京邮电大学 一种云环境下收益驱动的大规模处理任务调度方法
US20140214742A1 (en) * 2013-01-30 2014-07-31 Harris Corporation Distributed multi-layer particle swarm optimization based cognitive network
CN104176115A (zh) * 2013-05-22 2014-12-03 株式会社捷太格特 动力转向装置
CN105151117A (zh) * 2015-08-28 2015-12-16 南京航空航天大学 一种电控液压助力转向系统及基于该系统的多目标优化方法
CN106585714A (zh) * 2016-12-12 2017-04-26 南京航空航天大学 一种多模式复合转向系统分类控制器及其控制方法
CN107600173A (zh) * 2017-09-20 2018-01-19 南京航空航天大学 一种汽车液压变传动比转向系统及其多目标优化方法
CN108984917A (zh) * 2018-07-20 2018-12-11 北京航空航天大学 大型飞机飞控作动系统智能设计与评价方法
CN110386191A (zh) * 2019-04-23 2019-10-29 南京航空航天大学 一种电液智能转向系统性能监测器及性能优化方法

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105667580B (zh) * 2016-03-22 2018-08-24 南京航空航天大学 一种基于模糊控制的线控转向系统及其控制方法
CN106428197B (zh) * 2016-11-15 2018-11-09 南京航空航天大学 基于多模式转向系统助力耦合器的控制器及控制方法
CN107025337B (zh) * 2017-03-22 2020-04-24 南京航空航天大学 基于细胞膜优化算法的汽车复合转向系统多目标优化方法
CN109017974B (zh) * 2018-07-02 2020-12-25 南京航空航天大学 具有主动转向功能的辅助转向系统及其控制方法

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103019822A (zh) * 2012-12-07 2013-04-03 北京邮电大学 一种云环境下收益驱动的大规模处理任务调度方法
US20140214742A1 (en) * 2013-01-30 2014-07-31 Harris Corporation Distributed multi-layer particle swarm optimization based cognitive network
CN104176115A (zh) * 2013-05-22 2014-12-03 株式会社捷太格特 动力转向装置
CN105151117A (zh) * 2015-08-28 2015-12-16 南京航空航天大学 一种电控液压助力转向系统及基于该系统的多目标优化方法
CN106585714A (zh) * 2016-12-12 2017-04-26 南京航空航天大学 一种多模式复合转向系统分类控制器及其控制方法
CN107600173A (zh) * 2017-09-20 2018-01-19 南京航空航天大学 一种汽车液压变传动比转向系统及其多目标优化方法
CN108984917A (zh) * 2018-07-20 2018-12-11 北京航空航天大学 大型飞机飞控作动系统智能设计与评价方法
CN110386191A (zh) * 2019-04-23 2019-10-29 南京航空航天大学 一种电液智能转向系统性能监测器及性能优化方法

Also Published As

Publication number Publication date
CN110386191B (zh) 2020-07-24
CN110386191A (zh) 2019-10-29

Similar Documents

Publication Publication Date Title
WO2020215659A1 (zh) 一种电液智能转向系统性能监测器及性能优化方法
Huang et al. A BP-PID controller-based multi-model control system for lateral stability of distributed drive electric vehicle
WO2021077727A1 (zh) 一种电动卡车线控转向系统及其网络不确定控制方法
CN107791773B (zh) 一种基于规定性能函数的整车主动悬架系统振动控制方法
CN105741637B (zh) 四轮轮毂电机电动汽车智能转向控制方法
CN111002840B (zh) 一种分布式驱动电动汽车的容错控制方法
CN111152834B (zh) 一种基于阿克曼转向修正的电动汽车电子差速控制方法
US20050197994A1 (en) Intelligent robust control system for motorcycle using soft computing optimizer
Peng et al. Active safety control of X-by-wire electric vehicles: A survey
CN111873985A (zh) 一种四轮驱动电动汽车的集成底盘控制方法
CN113635879B (zh) 一种车辆制动力分配方法
Deng et al. Torque vectoring algorithm based on mechanical elastic electric wheels with consideration of the stability and economy
CN112829766B (zh) 一种基于分布式驱动电动车辆的自适应路径跟踪方法
CN114379583A (zh) 一种基于神经网络动力学模型的自动驾驶车辆轨迹跟踪系统及方法
CN103407341A (zh) 主动悬架支持向量机广义逆复合控制器及其构造方法
CN109017804B (zh) 力矩分配控制器为车辆各轮毂电机分配驱动力矩的方法
CN114987537A (zh) 基于神经网络动力学的自动驾驶车辆道路自适应漂移控制系统与方法
CN113341994B (zh) 一种基于分段仿射系统最优控制的智能汽车路径跟踪控制方法
CN109017447B (zh) 无人驾驶车辆总的驱动力矩输出方法
Kumar et al. Power consumption modeling of wheeled mobile robots with multiple driving modes
CN112918458B (zh) 一种全工况下的智能线控底盘能耗预测优化方法
Xu et al. Integrated optimal control of distributed in-wheel motor drive electric vehicle in consideration of the stability and economy
Lu et al. Adaptive heading control strategy for unmanned ground vehicle with variable wheelbase based on robust-active disturbance rejection control
Zhang et al. Research on coordinated control of braking mode switching for enhancing switching quality considering communication delay and time-varying model parameters
Xu et al. Linear time-varying MPC vehicle trajectory tracking controller considering driving road surface factors

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19926485

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19926485

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 19926485

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205 DATED 11-05-2023)

122 Ep: pct application non-entry in european phase

Ref document number: 19926485

Country of ref document: EP

Kind code of ref document: A1