Efficiency Optimization And Soft Start Control Of Induction Machines With Light Load

It is demonstrated that 90% part of power electricity are consumed in IM in China every year, the 60% of which belongs to light load even none load occasions and most of which runs under rated conditions without precise speed index and speed regulating equipment. It will lead to IM's efficiency weakness and energy loss if such kind of situation last long. Oriented to large and medium scaled IM, direct start easily causes current surge for the resons of structure, which influence IM's life and bring impacts to power grid. Facing to these problems, VVCF (Variable Voltage Constant Frequency) and VVVF (Variable Voltage Variable Frequency)can transmit un-rated mode to rated-mode, limit start current, and be good to realize IM's energy saving and soft-start. VVVF is helpful to realizing precise speed regulating; on the contrary, it adds to the costs of driven system, brings in high order harmonics which results in IM's harmonics power loss and power grid's pollutions. As a result, VVCF has wider application forehead and more important realistic meanings with low cost and flexible applicability under present energy crisis. In order to control the indices of efficiency and power loss in light load conditions close to the ones in rated conditions and to improve dynamic performances in start transient state, grounded on some type of oil squirrel cage extractor with light loads, efficiency's optimization and soft-start control issues in VVCF mode are studied in this thesis. The main contents include below:First, the voltage regulating platform is established based on new phase control techniques with optimized snubber circuit's parameters. VVCF is implemented through phase control techniques. In the studies of 3-phase anti-parallel connected thyristor voltage regulators, the bad effects'reasons for open-loop methods of firing angle and close-loop methods of holding-off angle are analyzed. On the basis of mechanical characteristics, power factor's variations, the controllable power factor angle is defined and a new phase control methods is provided based on driver output voltage complying with controllable power factor angle's variations. Simulation curves and experiment waveforms'contrasts of new methods and the previous two methods validate new methods'effects. Snubber circuit is adopted to inhabit driver's impacts to IM and to lower driver's power loss. Considering the factors of costs and practicability, passive resistance and capacitor snubber circuit's design and parameters'optimization are studied. The ideas are proposed of regarding impacts as constraints conditions, establishing energy loss index functions and minimizing the index functions to acquire snubber circuits'parameters. Two different methods of time-domain based on output characteristics and state-space based on electrical circuit modeling are employed to compute RC parameters. The effects of impacts'suppressions and energy-saving are validated by simulation curves and experiments waveforms.Secondly, in order to control the index of IM's power factor and efficiency in light load conditions close to the ones in rated conditions, a new slip frequency regulating method is supplied based on convex set optimizing and BP Artificial Neural Network (ANN) approaching. On the premise of known IM parameters and load ratios, power factor and efficiency's optimization can be taken as to solve multi-objectives optimizations with equation constraints conditions. Taking parameters'perturbations into considerations, the convex function relationship of index functions, optimized variables and IM's parameters are proved. The method based on ANN approaching prevents other rigrid methods from error. Taking stator resistance's temperature rise perturbation for example, simulation tests and experiments are done by means of BP networks'off-line training and regulator's on-line adjusting. Neural network's full excitation and close loop regulator add robustness to the system. Experiment waveforms validate results of multi-objectives optimization.Thirdly, in order to prevent control algorithm from control value'error and IM's parameters'perturbations and to balance the demands of smooth output torque and overload current, a new soft-start algorithm is raised based on BP ANN's nonlinear mapping. Face to the solutions'analysis of IM's dynamic equations under the d-q coordinates, neglecting IM's other nonlinear factors, driver output voltage and rotator frequency are the factors which influence soft-start transient state's dynamic performances. Oriented to the factors of output voltage, the experiment data based on controllable power factor methods in chapter 2 train the network in off-line mode to simulate the nonlinear feature of IM's soft-start. The new idea is put forward of exciting network by expected output values, the output of which being the needed control values. Simulation results release the effects of stimulus function and weight's variations to controllable values. Experiments based on new soft start algorithm are tested on the platform established in chapter 2 to verify its effectiveness and practicability.Finally, oriented to the other factor of rotator frequency proposed in chapter 4, a new controller containing rotator frequency's information is built on the basis of Lyapunov function. Taking holding-off angle negative feedback algorithm for example, by means of time-domain's modes analysis, frequency spectrum description, and open-circuit voltage and stator current's calculations, the principles of rotator frequency are obtained. The idea is proposed that controllable values should have the information of rotator's frequency to compensate the vibrations modes and to keep in power loss in transient state. Based on simulation experiment of hold-off angle negative feedback algorithm, simulation data of IM's open-circuit voltage and stator current during the course of soft start are acquired to derive rotator frequency'data and to establish speed observer. Based on the speed information output by speed observer, Lyapunov functions containing speed information with the quantities of energy are established and the voltage controller is designed and its has global stabilities. Comparing with the constant controller's results, the new IM controller's validities are verified.A software simulation platform based on MATLAB?\Simulink? is built for studying the provided control algorithms of VVCF controlled IM. A uniform IM simulation model invoking the resources of SimPowerSystems toolbox is established. In order to validate theoretical analysis, the experiment equipments based on MCS-51 series MCU controlled thyristor driven IM system is established. On the basis of simulation results, the relative validating experiments are done on the equipments...