Aimed at the problems of unbalanced capacitor neutral-point voltage and high common-mode voltage in a permanent magnet synchronous motor drive system powered by a T-type three-level inverter, a model predictive instantaneous torque control (MPITC) strategy based on finite voltage vector set optimization is proposed. First, in view of the influence of voltage vectors on the neutral-point voltage, only zero, small and large vectors are selected to participate in MPITC. Second, according to the relationship between the switching states and common-mode voltage, it is confirmed that 13 low common-mode voltage vectors participate in the control. To improve the operation performance of the motor, the long vector synthetic virtual voltage vector is used to replace the medium vector to participate in the model predictive control. Finally, according to the neutral-point potential and the motor current direction, the voltage vector which is favorable for maintaining the neutral-point voltage balance is selected from 19 voltage vectors as the preselected vector set. Experimental results show that the proposed control strategy can effectively reduce the electromagnetic torque, flux linkage pulsation and common-mode voltage amplitude, and the neutral-point voltage achieves balanced control.

To ensure the motor drive performance while saving cost in the field of electric industrial vehicles, low-precision encoders are often adopted. Although these encoders can measure the rotor’s accurate mechanical position, they introduce a long delay and large errors to the measured speed information. Therefore, the traditional load torque observers cannot obtain the accurate load torque information, and when the inaccurate information is used as feedforward of the current loop, the motor speed fluctuation cannot be effectively suppressed. To solve this problem, a second-order sliding-mode load torque observer based on a low-precision encoder is proposed, which can obtain the accurate load torque information based on the rotor position information and thus improve the anti-interference performance of the system. Finally, the correctness and effectiveness of the proposed method was verified by simulation and experimental results.

The development of industry and economy has caused a huge consumption of energy, which brings serious energy crisis and environmental pollution. Therefore, building a safe and clean energy interconnection network is a way to solve the relationship among social development, environment and energy at present. Nowadays, different countries have proposed their policies for the development of new energy electric vehicles (EVs). As the core component of EVs, lithium-ion batteries are directly related to the driving performance and safety of EVs. The state-of-charge (SOC) estimation is a core parameter of lithium-ion batteries used in various industries, and the estimation accuracy is directly related to the service life and efficiency of batteries. In this paper, the problem of battery SOC estimation accuracy in EV applications is studied, and an SOC estimation method based on the extended Kalman filter (EKF) optimized by the whale optimization algorithm (WOA) is proposed. On the basis of constructing the covariance matrix of system noise and observation noise, the improved and optimized WOA-EKF algorithm is used to optimize the noise covariance matrix under dynamic conditions, thus improving the SOC estimation accuracy. The model parameter identification and comparative simulation verification are carried out in MATLAB/ Simulink. Results show that the SOC estimation of lithium-ion batteries based on the WOA optimized EKF algorithm can control the SOC estimation error to be within 2% under different working conditions, which is of significance to the promotion of develop- ment of batteries in the new energy field.

Electric vehicle fast charging piles are prone to overheating of power devices under high-power operation, causing potential safety hazards. However, the existing cooling strategy adopts a rule-based forced air cooling method, and the cooling fan rotates at a high speed and generates large environmental noise. To protect the thermal safety of core components in the module while optimizing the cooling regulation strategy, an optimal thermal management method for electric vehicle fast charging module based on data-driven model predictive control (MPC) is proposed. This method adopts a data-driven method to construct a prediction model of module temperature distribution based on the long short-term memory neural network, and it combines MPC to control the fan speed, thus optimizing the thermal management strategy for the fast charging module and reducing the fan noise. Through experimental tests, it was verified that this method can effectively reduce the average fan speed by 1 293 rpm and reduce the average noise by 4.99 dB while ensuring that the key components are not overheated, which ensures the thermal safety of core components and the durability of the cooling fan.

A fast and accurate estimation of the state-of-charge (SOC) of lithium batteries is critical for the battery management system. Aimed at the problem that the Kalman filter algorithm lacks reasonable constraints on the resistance-capacitance (RC) parameters when estimating the SOC of lithium batteries, an optimization method of RC parameters filtering is proposed, and it is combined with unscented Kalman filter (UKF) to achieve the fast and accurate convergence of lithium battery SOC estimation. First, an equivalent circuit model of lithium battery is established by combing the polynomial equation. Then, forgetting factor recursive least squares is used to obtain the time-varying and time-invariant model RC parameters. The expression of RC parameters filtering relationship is established by setting the Kalman gain threshold, and an RC optimization UKF algorithm is proposed for lithium battery SOC estimation. Finally, hybrid pulse-power characteristic experiment, intermittent constant-current discharge experiment and dynamic stress test experiment were designed to verify the convergence and robustness of the proposed algorithm. The maximum estimation error of SOC was less than 1.0%, and the reference range of gain threshold was also given.

The accurate estimation of the state-of-charge (SOC) and state-of-health (SOH) of lithium-ion batteries is always a key scientific prob-lem that needs to be solved urgently. In this paper, based on a second-order fractional-order equivalent circuit model, the state space equation of a lithium-ion battery is established, and the discretization expressions of fractional-order differential and integral equations of battery parameters and SOC are derived. Then, a dual fractional-order extended Kalman filter method is studied to estimate the equivalent circuit parameters, SOC and battery capacity simultaneously. In addition, a time weighting sequence method based on estimated SOC and battery capacity is proposed, different discharge currents and cumulative time are monitored, and the available capacity of the battery is calculated online, thus achieving real-time estimation of the SOH of the battery at any discharge depth and any discharge rate. Finally, under the conditions of dynamic stress test, three lithium iron phosphate batteries of the same manufacturer, the same model and different aging degrees were used for experimental verification.

To enhance the accuracy of a lithium battery model and realize accurate state estimation of the lithium battery, a fractional-order electric model of the lithium battery was built, and the model parameters were identified using adaptive genetic algorithm. Based on the fractional-order electric model, the unscented Kalman filter was adopted to estimate the lithium battery’s state-of-charge (SOC) and state-of-health (SOH). Stimulation results show that, the established fractional- order electric model of the lithium battery can more accurately describe its dynamic characteristics during charging and discharging, and the accuracy of the proposed state estimation strategy was higher than that the conventional control strategy.

The switching speed and switching frequency of converters keep increasing, and higher requirements are also imposed on the DC-side decoupling capacitors. To solve the problems of overvoltage spikes and decoupling capacitor losses caused by voltage and current oscillations during the switch-off process, a transient circuit model considering the system’s stray parameters is established, and the evolution of overvoltage and decoupling capacitor current oscillation is analyzed. On this basis, a loss model considering the parasitic parameters of the main circuit and the equivalent series resistance of the decoupling capacitor is proposed. The limiting factors for the selection of decoupling capacitors in practical engineering are quantitatively analyzed, and an optimal selection method for the decoupling capacitance and loss limit condition is obtained. Finally, the proposed model and analysis method were verified by simulation and experimental results.

To improve the accuracy and efficiency of diagnosis when a converter fails, an active current limiting method is applied to realize the converter fault diagnosis in a power system. Active current limiting control is used to limit the output DC current from the converter to 1.2 times of rated current and limit the fault current, thus improving the stability of fault diagnosis. Based on a prediction model, the fault current distribution characteristics of the converter in the power system are diagnosed, and the bridge arm current is taken as the diagnosis parameter. The difference between the measured bridge arm current and predicted value is compared. If the difference is greater than the threshold value, it is judged that the converter has a fault. Experimental results show that the proposed method had a high accuracy and a high diagnosis efficiency in diagnosing the converter faults in an experimental power system, and the rates of false diagnosis and missed diagnosis were low, indicating that the converter diagnosis effect was satisfying.

In view of the recent situation in which power sources gradually reach their service terms in China, an evaluation method based on improved fuzzy analytical hierarchy process (FAHP) and entropy weight method (EWM) is proposed. First, an appropriate hierarchical structure is constructed based on the analytic hierarchy process to form a judgment matrix. Then, FAHP is used to process the judgment matrix between various layers, and the importance-oriented weight vector is obtained. At the same time, the Delphi survey method is used to form an evaluation matrix for the last sub criterion layer and the target layer. After normalization, EWM is used to obtain the value-oriented weight vector. The two weight vectors are synthesized to form a comprehensive weight vector. Finally, the final weight vector of the scheme layer to the target layer is formed, and the best scheme is given. The result of an example shows that the proposed evaluation method has strong flexibility and wide applicability. In addition, it also has a clear and reasonable process, as well as intuitive and accurate results.