Jihong Zhu

Adaptive Compensated Dynamic Inversion Control for a Helicopter with Approximate Mathematical Model

This paper presents an effective method to achieve altitude and attitude (pitch, roll and yaw) control of a helicopter in hovering and low-speed forward flight conditions with an approximate mathematical model. To facilitate control design, a simplified multi-input- multi-output (MIMO) affine nonlinear model that describes the angular rate responses of a helicopter is derived. The controller consists of two separated parts: altitude loop and attitude loop. A PID controller is used in the altitude loop and a dynamic inversion controller is used in the attitude loop. To compensate the dynamic inversion error caused by modeling uncertainties and disturbances, an adaptive compensating algorithm is employed. Simulation results of a YAMAHA R-50 unmanned helicopter simulation model show that the algorithm is stable and robust with good tracking performance and decoupling capability.

A Survey of Advanced Flight Control Theory and Application

Modern flight control system is a MIMO and cross-coupling nonlinear system, but traditional proportional-plus-integral strategy, with gain scheduling, optimizes parameters by linear feedback in one-loop-at-a-time and is hard to meet these rigid requirements, so it is emergent to explore advanced control approaches including linearity and nonlinearity today. Although recently some advanced flight control techniques have been demonstrated in modern aircrafts such as F-16, there are much more extensive problems to be studied such as system nonlinearity, deeply coupling between longitudinal- and lateralmotion, efficient evaluation of control redundancy allocation, flight path planning in complex environment, and on-line fault detection and self-repairing etc. This paper mainly discusses about modern flight control theory, design approaches and their applications. Author briefly introduces design approaches of modern flight control including linear control of linear quadratic regulator, robust flight control and nonlinear control of dynamic inversion and intelligent flight control etc. and shows their potential strengths and weaknesses, and discusses some promising applications of them as well

A novel path planning approach based on AppART and particle swarm optimization

Due to the NP-hard complexity, the path planning problem may perhaps best be resolved by stochastically searching for an acceptable solution rather than using a complete search to find the guaranteed best solution. Most other evolutionary path planners tend to produce jagged paths consisting of a set of nodes connected by line segments. This paper presents a novel path planning approach based on AppART and Particle Swarm Optimization (PSO). AppART is a neural model multidimensional function approximator, while PSO is a promising evolutionary algorithm. This path planning approach combines neural and evolutionary computing in order to evolve smooth motion paths quickly. In our simulation experiments, some complicated path-planning environments were tested, the result show that the hybrid approach is an effective path planner which outperforms many existing methods.

Nonlinear adaptive internal model control using neural networks for tilt rotor aircraft platform

An adaptive internal model controller using neural networks is designed for a tilt rotor aircraft platform. The behavior of the research platform, in certain aspects, resembles that of a tilt rotor aircraft. The proposed control architecture can alleviate the requirement of extensive gain scheduling of tilt rotor aircraft and compensate external disturbances, as well as dynamic inversion error. The controller includes an online learning neural network of inverse model and an offline trained neural network of forward model. Lyapunov stability analysis guarantees tracking errors and network parameters are bounded. The performance of the controller is demonstrated using the tilt rotor aircraft platform, with consistent response outcomes throughout experimental performing, including two nacelles tilting flight.

Composite dynamic surface control of hypersonic flight dynamics using neural networks

This paper addresses the composite neural tracking control for the longitudinal dynamics of hypersonicflight dynamics. The dynamics is decoupled into velocity subsystem, altitude subsystem, and attitudesubsystem. For the altitude subsystem, the reference command of flight path angle is derived for the attitudesubsystem. To deal with the system uncertainty and provide efficient neural learning, the composite law forneural weights updating is studied with both tracking error and modeling error. The uniformly ultimate boundednessstability is guaranteed via Lyapunov approach. Under the dynamic surface control with novel neuraldesign, the neural system converges in a faster mode and better tracking performance is obtained. Simulationresults are presented to show the effectiveness of the design.

Control allocation for a V/STOL aircraft based on robust fuzzy control

This study addresses the design of an output-feedback H-infinity and fuzzy mixed controller for an advanced vertical and/or short take-off and landing aircraft that exhibits satisfactory global stability criteria. A fuzzy logic based optimization approach is employed to solve autonomously the constrained control allocation problem by suitably adjusting the components of the output vector and finding a proper vector in the attainable moment set. Simulations show that the designed fuzzy controller is better suited to resist external disturbances, and is robust against parameter perturbations.

Eigenstructure Assignment Based Flight Control for Advanced Fighter: An Optimization Based Approach

An intelligent optimization approach is proposed for eigenstructure assignment (EA) via neural network (NN) adjusting the components of output vector autonomously. The basic idea is to minimize the L2 norm of error between the desired vector and achievable vector using the designing freedom provided by EA technique. Besides, close-loop eigenvalues are also optimised within desired regions on the left-half complex plane according to the design objective to ensure both closed-loop stability and dynamical performance. With the proposed approach, additional closed-loop specifications such as decoupling of different modes and robustness can also be easily achieved. As a demonstration, application of the proposed approach to the designing of flight control law for an advanced fighter is discussed. The simulation results show good closed loop performance and validate the proposed intelligent optimization approach of EA technique.

Robust adaptive neural control of a class of nonlinear systems

In this paper we present a robust adaptive control for a class of uncertain continuous time multiple input multiple output (MIMO) nonlinear systems. Multiple multi-layer neural networks are employed to approximate the uncertainty of the nonlinear functions, and robustifying control terms are used to compensate for approximation errors. All parameter adaptive laws and robustifying control terms are derived based on Lyapunov stability analysis so that, under appropriate assumptions, semi-global stability of the closed-loop system is guaranteed, and the tracking error asymptotically converges to zero. Simulations performed on a two-link robot manipulator illustrate the approach and its performance.

Letters: Experimental modeling using modified cascade correlation RBF networks for a four DOF tilt rotor aircraft platform

Modeling using the modified cascade correlation radial basis function (CCRBF) networks for an experimental platform is presented. We develop a practical algorithm coupled with model validity tests for identifying nonlinear autoregressive moving average model with exogenous inputs (NARMAX). The effectiveness of this modeling procedure is demonstrated by a four degrees-of-freedom (DOF) platform, and the experimental results show that the modified CCRBF networks for the platform are more appropriate than the conventional ones. The estimated model can be utilized for nonlinear flight simulation and control studies.

Genetic Algorithm Based Constrained Control Allocation for Tailless Fighter

A genetic algorithm based optimization approach is proposed for constrained control allocation problem via adjusting the components of output vector and finding a proper vector in the attainable moment set (AMS) autonomously. The basic idea is to minimize the L2 norm of error between the desired moment and attainable moment using the designing freedom provided by redundant control surfaces. With the constraints of control surfaces, in order to obtain desired performance of aircraft such as stability and maneuverability, the weights of different components are updated by the genetic algorithm, which makes the close-loop system self-adaptation. As a demonstration, application of the proposed approach to the designing of control system for a tailless fighter is discussed. The results show good closed loop performance and validate the proposed intelligent optimization approach of constrained control allocation for flight control