Showing papers presented at "International Power Electronics and Motion Control Conference in 1998"

Modelling and implementation of a permanent magnet synchronous motor drive using a DSP development environment

Abstract: A permanent magnet synchronous motor drive using a prototype 6 pole 3 kW inset magnet motor is presented. The drive is controlled by a DSP based control board that can be programmed using the MATLAB/SIMULINK software. To investigate the influence of saturation, an accurate measurement procedure based on load tests is described. This method benefits from the I/O features of the development environment. The theoretical performance limits of the motor are determined using optimal torque control. Finally, aspects of controller design, simulation and implementation are briefly discussed.

Trajectory Tracking with SMC Technique under Load Disturbance: Experimental Results

Abstract: Tlus paper deals with the application of a sliding mode control (SMC) in tracking systems with trajectory generation. The possibilities of using this kind of VSS strategy on tracking control with reference trajectory generation under load disturbance and variable inertia is tested. The practical design of such robust sliding mode controller is discussed. Control law is obtained by using the methodology of balance condition. In order to achieve a tracking with quaranteed precision, smoothing of control function in constant and variable boundary layer are compared. Laboratory setup for experimental verification are realised. It consists of host system (based on MC 68332) and tw mechanically coupled synchronous motor permanent magnet (SMPM), where one motor is load for another one. Economical loading is used, connecting the DC Busbars of both converters together so only a minimum energy for losses are consummated. K e w r d s . Sliding mode control, robust control, balance condition, tracking system, trajectory generation, economical loading. LNTRODUCTION E~ E =xi , -k,,&/k, =x2, The sliding mode is a special case of Variable Structure Systems (VSS) and keeps invariant trajectory of moving under different plants uncertainties. It is especially suited for systems where robustness is a crucial performance requirement. Many practical applications [4], [5], [6] confirmed that the robusi nature is quaranteed by sliding mode. This nature is achieved with control algorithm very simple and easy to implement in real time computer control systems. The main dra*ack of sliding mode is that the resulting control input is discontinuous on the switching surface and, consequently, the control input chatters at a theoretically infinite frequency. Chattering is highly undesirable, since it involves extremely high control activity, and furthermore may excite hgh-frequency dynamics neglected in the course of modelling. To overcome this problem, the discontiriuous function is replaced by a proper control function which consists of continuos part, i.e. equivalent control [l], 131, [7], and discontinuous part (relay type component). In addition, discontinuous part in control input is replaced by continuos one in t hn boundary layer. According to balance condition, boundary layer thickness can be made time-varying [2], [3], [4], [6]. In that case one can spec* the best attainable tracking performance, given the desired control bandwidth and the extent of load variations and parameter uncertainty. and eliminating x2 we can show that is: where E is actual position, u(t) control input, m,(t) load torque, G,,=k,/(l+yTzJ = k:, transfer function i,(y)/i,'(y) , kc coefficient in position feedback loop, k ~ " D/A converter constant, k, torque constant and J inertia of the drive. Taking that k,=kDA k:, k, k,N and k-=k,/J . we have from ( I ) and (2): E-' =kl u(t)-k2m, ( t ) . (3) With F = E E ~ and ; = i. -id sliding function is defined as [2]: where F , .cd are error position and desired position. F , Ed are error speed and desired speed. For continuos part of control input based on equivalent control (u,,, s o ) , using (3) and (4). we find the best approximation ti for equivalent control: where i,i1 ,i, ,h, are estimated values of continuos DESIGN OF SLIDING MODE CONTROLLER part (equivalent control) u,,, coefficients k,, k, and load torque m, respectively. Then, we find the total For the system in Fig. 1. we find that is: control input as: