A discrete-time active disturbance rejection control (ADRC) method using resonant sliding mode is proposed to solve the disturbance characteristics of the current loop of permanent magnet (PM) motor at low speed. The extended state observer (ESO) is used to estimate the disturbance, but there is a contradiction between disturbance rejection and noise rejection in parameter design of ESO, which will affect the accuracy of the estimation, especially at low speed. In order to solve the problem, sliding mode control is combined with ADRC, while traditional integral sliding mode control still has poor rejection effect on the specific frequency harmonics. The resonant sliding mode is proposed to add resonance term to the integral sliding mode surface, which mitigates the impact of the disturbance that ESO fails to estimate on the tracking accuracy. Therefore, resonant sliding mode in the feedback control law of ADRC (RSM-ADRC) is employed to current loop to reduce the steady-state error. Meanwhile, the stability of the proposed algorithm is verified by discretization method. The steady and dynamic performances of the four control methods are compared through experiments to verify the feasibility and effectiveness of the proposed method. For example, when the reference current of d - and q -axis is 0A and 1A respectively at 50rpm, the total harmonic distortion of q -axis current decreases from 1.68% to 0.64% and the 6th harmonic of current also drops to 0.135%. Despite the sudden change of inductance parameters, the steady-state current response is largely unaffected.
This paper proposes a double closed-loop predictive control scheme for a permanent-magnet synchronous motor system, which consists of a speed loop model predictive control and a current loop deadbeat predictive control, current loop and speed loop adopt different predictive control methods, compensate the unstable switching frequency, and reduce the amount of calculation. Considering the influence of parameter mismatch on the control performance, a parameter disturbance sliding mode observer and a parameter disturbance internal mode observer are designed to compare the compensation ability of the two observers for current tracking errors, which improves the disturbance rejection ability and robustness of the system, and the compensation ability of the internal mode observer is better. The simulated and experimental results show that the improved strategy and predictive control method proposed in this paper can obtain better speed tracking performance compared with traditional control, which verify the effectiveness and feasibility of the proposed control scheme on a servo control system.
In order to ascertain the impact of the Tohoku-Oki 3.11 M9.0 earthquake on the stability of the faults in the Tangshan seismic region, we investigated the adjustment of the in situ stress field of the seismic region after this earthquake based on in situ stress monitoring data. Then, according to the Mohr-Coulomb failure criteria and Byerlee’s law, we used the FSP v.1.0 software package to calculate the fault slip potential (FSP) of the main faults in the seismic region at each adjustment stage of the in situ stress field, and to study the risk of fault activity. The research results show that 1) after the Tohoku-Oki 3.11 M9.0 earthquake, the tectonic environment of the Beijing Plain area changed rapidly from nearly EW extrusion to nearly EW extension, and this state was maintained until June 2012. After this, it began to gradually adjust to the state present before the earthquake. As of September 2019, at the depth of 100m, the maximum horizontal principal stress value in Tangshan seismic region was 7.61–7.81MPa, the minimum horizontal principal stress value was 5.30–5.50MPa, and the maximum horizontal principal stress orientation was N55.1°–59.5°E. (2) Before the Tohoku-Oki 3.11 M9.0 earthquake, the stress accumulation level of the main faults in the seismic region was relatively high with the FSP values of 30–60%. After this earthquake, the stress accumulation level of each fault continued to decrease, as of May 2013, the FSP values were mainly concentrated at 10–35%. Then, the stress accumulation level of major faults in the seismic region began to gradually increase. As of September 2019, the FSP values were mainly concentrated at 23–37%, and the stress accumulation level was still lower than the pre-earthquake state. 3) The fault activity in the central and northern parts of the seismic region was the strongest, followed by the southern part and western part, and the fault activity in the eastern part was the weakest.