Written in two parts, the first revises the ideas and theoretical bases necessary for a good understanding of the techniques used in the second, which deals with applications of MATLAB(R) and SIMULINK(R) in process control and digital signal processing. Each application is treated through various techniques including the classical methods of automation and of deterministic and random digital processing using fuzzy logic and neural networks. The preceding mathematical study of the physical processes goes from finding the equations to editing the analogical model. The following SIMULINK(R) toolbox functions and blocks have been used: Control System, Signal Processing, Neural Network and Fuzzy Logic.

Inhalt

1 : Analog and digital control.- 1. The principle.- 2. Presentation of main types of corrector.- 2.1. Proportionnal corrector.- 2.2. Integral control.- 2.3. Derivative corrector.- 2.4. Derivative return corrector.- 2.5. Phase lead corrector.- 2.6. Phase lag corrector.- 2.7. PID controller.- 2.8. Predictive action corrector.- 2.9. PIR corrector, pure delay system.- 3. Analog correctors discretisation.- 4. Corrected systems stability.- 4.1. General conditions of stability.- 4.2. Nyquist criterion.- 4.3. Discrete systems stability.- 5. Examples.- 5.1. Using some MATLAB® functions.- 5.2. Using a PIR corrector.- 6. LQ, LQI, quadratic linear control.- 6.1. LQI control of a monovariable process.- 6.1.1. Model without integrator.- 6.1.2. Model with integrator.- 6.2. LQI control of a multivariable process.- 6.2.1. LQ multivariable control.- 6.2.1. LQI multivariable control.- 6.3. Application example.- 6.3.1. LQI control of a monovariable aerothermal system.- 6.3.2. LQI control of a multivariable system.- 7. RST control.- 7.1 Monovariable system.- 7.2 Multivariable system.- 7.3 Application example.- 7.3.1. RST monovariable control of the temperature.- 7.3.2. RST multivariable control of the aerothermal process.- 2: State representation of continuous and discrete systems.- 1. State representation of continuous systems.- 1.1. Heuristic approach.- 1.2. State representation generalization.- 2. State representation of discrete systems.- 3.1. Heuristic approach.- 3.2. Application.- 3. Controllability and observability.- 3.1. Controllability.- 3.2. Observability.- 4. State reconstruction of a discrete dynamic system.- 4.1. Closed-loop estimation of a deterministic process.- 5. State return control.- 6. Examples.- 6.1. State return control system of a process including an integration.- 6.2. State return control system of a process not including an integration.- 6.3. Control system by poles placing of a discrete system.- 7. Kalman filter.- 8. Discrete stochastic Kalman predictor.- 3: Fuzzy logic control.- 1. The fundamental principle.- 2. Stages of implementation of a fuzzy regulator.- 2.1. Fuzzification stage.- 2.2. Inference stage.- 2.3. Defuzzification stage.- 3. Graphical interface of the "Fuzzy Logic TOOLBOX".- 4. Creation of a fuzzy system using the toolbox commands.- 4.1. Input and output variables fuzzification.- 4.2. Fuzzy Rules Editor.- 4.3. Defuzzification.- 4.4. Using the regulator in a control law.- 5. Fuzzy regulator use in SIMULINK®.- 6. Sugeno's method.- 6.1. Realisation of the fuzzy regulator using the graphic interface.- 6.1. Realisation of the fuzzy regulator using the TOOLBOX commands.- 4: Neural networks.- 1. Introduction.- 2. Linear adaptive neural networks.- 2.1. Architecture.- 2.2. Training law.- 2.3. Some applications fields.- 2.3.1. Process identification.- 2.3.2. Signal prediction.- 2.3.3. Interference cancellation.- 3. Neural networks with hidden layers, back-propagation error.- 3.1. Principle.- 3.2. Transfer functions.- 3.3. Back-propagation algorithm.- 4. Inverse neural model control.- 4.1. First architecture.- 4.2. Second architecture.- 4.2.1. Addition of an integration.- 4.2.2. Adaptive control.- 5. Signal prediction.- 5: Adaptive filtering.- 1. The adaptive filtering principle.- 2. Gradient algorithm, LMS criterion.- 2.1. ? scalar adaptation choice.- 2.2. Adaptation speed, filter time constant.- 3. The recursive least squares algorithm, exact least squares criterion.- 4. Examples of LMS adaptive filters.- 4.1. Adaptive predictor for an autoregressive process.- 4.1. Interference cancellation.- 4.1. Extraction of a signal drowned in noise.- 5. RLS adaptive filter example.- 5.1. Extraction of a signal drowned in noise.- Application 1: Power amplifier.- 1. Description of amplifier.- 2. Characterization of amplifier.- 3. Amplifier with transistors stage feedback.- 4. Amplifier with phase lag corrector.- 5. Amplifier with feedback of phase lead type corrector.- Application 2: Electromagnetic levitation.- 1. Process modelling.- 1.1. Expression of the F attraction power according to the I current in the coil and the e air gap.- 1.2. Process linearization around a quiescent point e(t)=e0.- 1.3. Process transfer functions.- 2. Electric current amplifier control system.- 3. Analogical and discrete models of the x(t) position control system.- 4. x(t) digital follower control.- 5. Using a fuzzy corrector.- 5.1. Variables fuzzification.- 5.2. Inference rules definition.- 5.3. Output defuzzification.- Application 3: Cart with inverted pendulum.- 1. System modelling with 2 degrees of freedom.- 1.1. Kinetic energy of the system on motion.- 1.2. Potential energy of the system.- 1.3. Lagrange equation according to q(t)=?(t) degree of freedom.- 1.4. Lagrange equation according to q(t)=x(t) degree of freedom.- 1.5. Linear model around the operating point.- 2. Linear process state modelling.- 3. Edition and test of the discrete model.- 4. Fuzzy regulation of the ?(t) angular position.- 4.1. Inputs fuzzification, membership functions definition.- 4.2. Inference rules definition, defuzzification.- 4.3. Achieving the fuzzy controller.- 5. Fuzzy control of the x(t) position and the ?(t) angle.- 5.1. Inputs fuzzification, membership functions.- 5.2. Inference rules definition, defuzzification.- 5.3. Achieving the fuzzy controller.- 6. Graphical animation of the system.- Application 4: Oven control.- 1. Oven modelling.- 2. Integral control with compensation of poles and zeros.- 3. Discrete state representation of the oven.- 4. Control by state return with integration.- 5. Using a Kaiman reconstructor.- 6. LQ quadratic linear control.- 7. Control by neuronal inverse model.- Application 5: Travelling gantry crane with suspended mass.- 1. Modelling the travelling gantry crane with 2 degrees of freedom.- 1.1. Kinetic energy of the system on motion.- 1.2. Potential energy of the system.- 1.3. Lagrange equation for the q(t)=?(t) degree of freedom.- 1.4. Lagrange equation for the q(t)=x(t) degree of freedom.- 1.5. Linear model upon the operating point.- 2. Transfer functions of the system.- 2.1. Step response of the open loop process.- 2.2. Edition and test of the model.- 3. Regulation of the ?(t) angular position.- 4. Regulation of the x(t) position truck and the ?(t) angle.- 5. State space modelling.- 5.1. Discrete state space model.- 5.2. Luenberger's state observer.- 5.3. State space control of the process.- 5.4. Adding an integral correction.- 6. Graphical animation of the travelling gantry crane.- 7. Fuzzy control of the gantry.- 8. RST and LQI controllers.- 8.1 Discrete model of the gantry.- 8.2 RST control law.- 8.2.1. RST monovariable control of the truck position.- 8.2.2. Multivariable RST control of the travelling gantry crane.- 8.3. LQI mono variable control of the cart position.- Application 6: Hands-free telephone.- 1. Programming Adaline using MATLAB® commands.- 2. Using S-function in a SIMULINK® model.- Application 7: Echo cancellation on a transmission line.- 1. Transmission line modelling.- 2. LMS filtering, lms1 S-function.- 3. RLS filtering, rlsl S-function.- Application 8 : Noise elimination in a conduit.- 1. Conduit modelling.- 2. LMS filtering, lms2 S-function.- 3. RLS filtering, rls2 S-function.- 4. Composite noise filtering.- Ap…