Application of Learning Automata to Controller Design in Slow-Active Automobile Suspensions

SUMMARY

This study considers a new design methodology in the context of active vehicle suspension control. The approach combines concepts from Stochastic Optimal Control with those of Learning Automata. A learning automaton effectively learns optimal control on-line in the vehicle, in an appropriate stochastic “test-track” environment. For practical application, the overwhelming advantage of this approach is that no explicit modelling is required, and considerable time savings may be expected in system development. This simulation study considers the on-line learning of optimal control in a low-bandwidth active suspension system, where control feedback is confined to a body-mounted accelerometer at each corner of the vehicle. It is shown that learning can successfully take place under a range of conditions, including the case when there is substantial transducer noise. The performance of the resulting control system is shown to depend heavily on the nature of the learning environment.     

Optimal Performance of Variable Component Suspensions

Most vehicle suspension systems use fixed passive components that offer a compromise in performance between sprung mass isolation, suspension travel, and tireroad contact force. Recently, systems with discretely adjustable dampers and air springs been added to production vehicles. Active and semi-active damping concepts for vehicle suspensions have also been studied theoretically and with physical prototypes. This paper examines the optimal performance comparisons of variable component suspensions, including active damping and full-state feedback, for “quartercar” heave models. Two and three dimensional optimizations are computed using performance indicators to find the component parameters (control gains) that provide “optimal” performance for statistically described roadway inputs. The effects of performance weighting and feedback configuration are examined. Active damping is shown to be mainly important for vehicle isolation. A passive vehicle suspension can control suspension travel and tire contact force nearly as well as a full state feedback control strategy.     

DSP-based self-organising fuzzy controller for active suspension systems

This study proposed a self-organising fuzzy controller (SOFC) for controlling an active suspension system to evaluate its control performance. During the control process, the SOFC continually updated the learning strategy in the form of fuzzy rules. The fuzzy rule table of this SOFC could be initially set to zero. This not only overcame the difficulty in finding appropriate membership functions and control rules for designing a fuzzy controller, but also solved the database problem where the fuzzy rules of a fuzzy controller, once determined, remained fixed and could not suitably regulate them in real time to optimise the dynamic response of the system required to gain the desired control performance. To demonstrate the applicability of the proposed SOFC for active suspension systems, a quarter-car hydraulic-servo suspension system was designed and constructed to evaluate the feasibility of active suspension control. Additionally, to conform to real-time application requirements in the vehicular industry, the SOFC was implemented with a digital signal processor to control the hydraulic-servo suspension system so that the control performance could be determined. The SOFC has shown a better control performance in suppressing the vibration amplitude of the vehicle body for enhancing the structural safety of the vehicle and increasing the life of the suspension system. It also effectively suppressed the amplitude of the vehicle body acceleration and reduced the tire deflection for improving the ride and the handling quality of a vehicle better than a passive control, as verified in experimental results.     

Performance of Low-bandwidth, Semi-active Damping Concepts for Suspension Control

Active damping has been shown to offer increased suspension performance in terms of vehicle isolation, suspension packaging, and road-tire contact force. It can even approximate the performance of full state feedback control without requiring the difficult measurement of tire deflection. Many semi-active damping strategies have been introduced to approximate the response of active damping with the modulation of passive damping parameters. These strategies have typically required a relatively high bandwidth for actuator response. This paper investigates the simulation performance and “frequency response” of two concepts in low-bandwidth semi-active suspension control, one that sets a damping force directly and another that sets the damping resistance. The electronically controlled bandwidth of these actuators is approximately an order of magnitude less than other semi-active devices; high frequency control is handled mechanically. A quarter-car model is studied with the controlled damping replacing both passive and active damping of typical control schemes. Both low-bandwidth damping strategies perform remarkably well compared to both active and high-bandwidth, semi-active damping. In certain dynamic performances, the new semi-active strategies outperform active damping and what the author calls “nominal” semi-active damping.     

Road disturbance estimation for the optimal preview control of an active suspension systems based on tracked vehicle model

This research investigates stochastic estimation of a look-ahead sensor scheme using the optimal preview control for an active suspension system of a full tracked vehicle (FTV). In this scheme, wheel disturbance input to the front wheels are estimated using the dynamic equations of the system. The estimated road disturbance input at the front wheels are utilized as preview information for the control of subsequently following wheels of FTV. The design of optimal preview control is used as a classical linear quadratic Gaussian problem by combining dynamics of the original system and estimation of previewed road inputs. The effectiveness of the preview controller is evaluated by comparing the estimated information with the measured information for different road profiles, where Kalman filter is used for the state-variables estimation of the FTV. This research also considers the reduced order estimation using commonly available sensors in order to decrease the number of sensors and measurements. The simulation results’ using an active suspension system with different preview information shows that the proposed system can be beneficial for the improvement of ride comfort of tracked vehicles without using any specialized sensors for preview information calculation.     

基于SIMULINK的车辆主动悬架LQG控制仿真研究

通过建立1／4车辆模型,应用最优控制理论进行了车辆主动悬架的LQG(Linear Quadratic Gaussian)控制器的设计,并在Matlab／Simulink环境中建立系统模型并进行仿真,将仿真结果与被动悬架仿真结果进行对比分析。仿真结果表明,具有LQG控制器的主动悬架对车辆行驶平顺性和乘坐舒适性的改善有良好的效果。     

Optimal Performance of Variable Component Suspensions

SUMMARY

Most vehicle suspension systems use fixed passive components that offer a compromise in performance between sprung mass isolation, suspension travel, and tireroad contact force. Recently, systems with discretely adjustable dampers and air springs been added to production vehicles. Active and semi-active damping concepts for vehicle suspensions have also been studied theoretically and with physical prototypes. This paper examines the optimal performance comparisons of variable component suspensions, including active damping and full-state feedback, for “quartercar” heave models. Two and three dimensional optimizations are computed using performance indicators to find the component parameters (control gains) that provide “optimal” performance for statistically described roadway inputs. The effects of performance weighting and feedback configuration are examined. Active damping is shown to be mainly important for vehicle isolation. A passive vehicle suspension can control suspension travel and tire contact force nearly as well as a full state feedback control strategy.     

Twenty First Century Transportation System Solutions - a New Type of Small, Relatively Tall and Narrow Active Tilting Commuter Vehicle

There are many problems that face transportation systems as the twenty-first century approaches, and many solutions will be required. Mass transportation systems are one large area of research that will provide some solutions. This paper presents another possible solution at the other end of the spectrum, small relatively tall and narrow tilting commuter vehicles for individual transportation. A historical overview of the various types of tilting vehicles built or proposed over the last forty years is shown and the results of these studies are discussed. If one considers a relatively tall and narrow vehicle, (under 1.0 meters or 40“ wide), to maintain high speed performance in cornering it becomes necessary to bank the vehicle into a corner to prevent overturning. The design of a modem active tilting suspension and control law for a small narrow, one-half width, commuter vehicles is presented. Analysis of the static and dynamic tipping limits illustrates which vehicles are considered tall and narrow requiring active tilting. The performance of such vehicles as they enter a steady corner is considered and how tilt dynamics may feel to passengers is discussed.     

Level and attitude control of the active suspension system with integral and derivative action

A 7-DOF full-car model with optimal active control suspension is utilized to evaluate the vehicle dynamic performances which are achieved through proposed controllers. The optimal controller, which includes the integral action for the suspension deflection, considerably improves the attitude control of a vehicle because the rolling and pitching motion in cornering and braking maneuvers are reduced, respectively. In the viewpoint of level control, the integral control acting on the suspension deflection results in the zero steady-state deflection in response to static body forces and ramp road input. The dynamic characteristics of the suspension control system are evaluated in terms of time domain and frequency domain. The simulations in the time domain demonstrate the advantages of the active suspension system obtained by penalizing the integral and derivative of suspension deflections and the derivative of roll and pitch angles in the performance index. The frequency characteristic curves obtained by simulations regarding integral action or derivative action show the increase of both ride comfort and road-holding performances by maximizing the use of suspension deflections. The potential of derivative control is shown by the performances of the car traveling over a bump and braking.     

Level and attitude control of the active suspension system with integral and derivative action

A 7-DOF full-car model with optimal active control suspension is utilized to evaluate the vehicle dynamic performances which are achieved through proposed controllers. The optimal controller, which includes the integral action for the suspension deflection, considerably improves the attitude control of a vehicle because the rolling and pitching motion in cornering and braking maneuvers are reduced, respectively. In the viewpoint of level control, the integral control acting on the suspension deflection results in the zero steady-state deflection in response to static body forces and ramp road input. The dynamic characteristics of the suspension control system are evaluated in terms of time domain and frequency domain. The simulations in the time domain demonstrate the advantages of the active suspension system obtained by penalizing the integral and derivative of suspension deflections and the derivative of roll and pitch angles in the performance index. The frequency characteristic curves obtained by simulations regarding integral action or derivative action show the increase of both ride comfort and road-holding performances by maximizing the use of suspension deflections. The potential of derivative control is shown by the performances of the car traveling over a bump and braking.     

Application of Nonlinear Control Theory to Electronically Controlled Suspensions

This paper illustrates the use of nonlinear control theory for designing electro-hydraulic active suspensions. A nonlinear, “sliding” control law is developed and compared with the linear control of a quarter-car active suspension system acting under the effects of coulomb friction. A comparison will also be made with a passive quarter-car suspension system. Simulation and experimental results show that nonlinear control performs better than PID control and improves the ride quality compared to a passive suspension.     

Stochastic Optimal Control of Highway Tractors with Active Suspensions

Stochastic optimal control and estimation theories are used to design an active suspension system for a cab ride in a tractor-semitrailer vehicle. A discrete-continuous vehicle model with eleven degrees of freedom is augmented by a stochastic road excitation model and a human perception of vibration shape filter. Both perfect measurement and estimated state cases are considered. The impact of the measurement noise on the design of the optimal controller is demonstrated. The performance of the optimally controlled system is compared with an optimal passive system. It is shown that significant improvements in ride comfort can be achieved through the use of actively controlled cab suspensions.     

Adaptive neuron control using an integrated error approach with application to active suspensions

This paper proposes a new neuron control strategy for an active vehicle suspension system, with the emphasis on the study of multivariable and uncertain suspension characteristics. The novelty of this strategy is in the use of integrated error, which consists of multiple output errors in the regulated plant. By combining the integrated error approach with the traditional neuron control (TNC), integrated error neuron control (IENC) is presented. It provides a direct control to the multiple outputs of the control plant simultaneously. Taking a quarter-car model as an example, the proposed control strategy is applied and comparative simulations are carried out with various vehicle parameters and road input conditions. Simulation results prove the effectiveness and robustness of the proposed IENC method. In addition, the newly proposed neuron scheme provides a simple yet efficient new possibility for the control of a class of uncertain multivariable systems similar to an active vehicle suspension.     

Optimal Linear Preview Control of Active Vehicle Suspension

SUMMARY

The problem of linear preview control of vehicle suspension is considered as a continuous time stochastic optimal control problem. In the proposed approach minimal a priori information about the road irregularities is assumed and measurement errors are taken into account. It is shown that estimation and control issues can be decoupled. The problem formulation and the analytical solution are given in a general form and hence they apply to other problems in which the system disturbances are unknown a priori, even in a stochastic sense, but some preview information is possible.

The solution is applied to a two-degree-of-freedom (2-DOF) vehicle model. The effects of preview information on ride comfort, road holding, working space of the suspension and power requirements are examined in time and frequency domains. The results show that the greatest potential is for improving road holding properties. This effect could not have been observed in previous studies based on a 1-DOF vehicle model. It is also demonstrated that the presence of preview drastically reduces power requirements, thus relieving the performance versus actuator power dilemma.     

Optimal Adaptive Active Suspensions for a Full Car Model

In order to present a useful method for designing active suspension of a vehicle, a linear full-car model is used in this investigation. In this model, the dampers of passive system are totally replaced by actuators. The actuators are controlled with optimal full state vector feedback. After determining feedback coefficients, the responses of active and passive systems were compared and it was found that performance of active system is much superior. It is desired that, changes in vehicle parameters would not affect the system's performance and hence should not violate its optimality. In other words, the system should behave adaptively using Model Reference Adaptive Control. The optimally controlled active suspension was used as a model for the active suspension of vehicle. In this way, the suspension of vehicle is controlled in such a way that its output approaches to that of the optimal active model. Thus the suspension should behave just like the optimal one.     

DISCUSSION NOTE on Karnopp's article “Active Damping in Road Vehicle Suspension Systems”

SUMMARY

In this paper, an optimal suspension system is derived for a quarter-car model using multivariable integral control. The suspension system features two parts. The first part is an integral control acting on suspension deflection to ensure zero steady-sate offset due to body and maneuvering forces as well as road inputs. The second is a proportional control operating on the vehicle system states for vibration control and performance improvement. The optimal ride performance of the active suspensions based on linear full-state feedback control laws with and without integral control together with the performance of passive suspensions are compared.     

Utilisation of optimisation solutions to control active suspension for decreased braking distance

This work deals with how to utilise active suspension on individual vehicle wheels in order to improve the vehicle performance during straight-line braking. Through numerical optimisation, solutions have been found as regards how active suspension should be controlled and coordinated with friction brakes to shorten the braking distance. The results show that, for the studied vehicle, the braking distance can be shortened by more than 1?m when braking from 100?km/h. The applicability of these results is studied by investigating the approach for different vehicle speeds and actuator stroke limitations. It is shown that substantial improvements in the braking distance can also be found for lower velocities, and that the actuator strokes are an important parameter. To investigate the potential of implementing these findings in a real vehicle, a validated detailed vehicle model equipped with active struts is analysed. Simplified control laws, appropriate for on-board implementation and based on knowledge of the optimised solution, are proposed and evaluated. The results show that substantial improvements of the braking ability, and thus safety, can be made using this simplified approach. Particle model simulations have been made to explain the underlying physical mechanisms and limitations of the approach. These results provide valuable guidance on how active suspension can be used to achieve significant improvements in vehicle performance with reasonable complexity and energy consumption.     

Development of Preview Active Suspension Control System and Performance Limit Analysis by Trajectory Optimization

The main role of the suspension system is to achieve ride comfort by reducing vibrations generated by the road roughness. The active damper is getting much attention due to its reduced cost and ability to enhance ride comfort especially when the road ahead is measurable by an environment sensor. In this study a preview active suspension control system was developed in order to improve ride comfort when the vehicle is passing over a speed bump. The control system consists of a feedback controller based on the skyhook logic and a feedforward controller for canceling out the road disturbance. The performance limit for the active suspension control system was computed via trajectory optimization to provide a measure against which to compare and validate the performance of the developed controller. The simulation results indicated that the controller of this study could enhance ride comfort significantly over the active suspension control system employing only the skyhook feedback control logic. Also the developed controller, by displaying similar control pattern as the trajectory optimization during significant time portions, proved that its control policy is legitimate.     

Active Suspension Control to Improve Vehicle Ride and Handling

In practice most active vehicle suspension work can be traced to two sources, Lotus' modal control and Karnopp's skyhook damper. A model is developed which allows comparison of different active suspension control algorithms. The Lotus modal control algorithm is reviewed, and compared with Karnopp's skyhook damper. It is shown that a tight inner closed loop allows the Lotus algorithm to achieve the inertial damping described by Kamopp for a single comer or quarter car. It is suggested that to achieve simultaneously high inertial damping and good disturbance rejection an inner force loop is desirable. A vehicle control scheme is presented which combines the Lotus modal decomposition with Karnopp's skyhook damper, allowing nearly optimal ride and simultaneously permitting modification of vehicle handling properties.     

Fuzzy and adaptive fuzzy control for the automobiles’ active suspension system

Two typical criteria for good vehicle suspension performance are their ability to provide good road handling and increased passenger comfort. The main disturbance affecting these two criteria is terrain irregularities. Active suspension control systems reduce these undesirable effects by isolating car body motion from vibrations at the wheels. This paper describes fuzzy and adaptive fuzzy control (AFC) schemes for the automobile active suspension system (ASS). The design objective is to provide smooth vertical motion so as to achieve the road holding and riding comfort over a wide range of road profiles. The efficacy of the proposed control schemes is demonstrated via simulations. With respect to the optimal linear quadratic regulator (LQR), it is shown that superior results have been achieved by the AFC.