A Nonlinear PI Controller for Speed Control of Electric Drives Using Radial-Basis Function Neural Network

In this paper, a nonlinear proportional and integral (PI) controller based on Radial Basis Function Neural Network (RBFNN) is proposed for the speed control of electric drives. The Lyapunov function is employed in the design process to ensure system stability. The proposed nonlinear PI controller has a fixed proportional gain but a variable integral gain, which makes it outperforms than the conventional linear PI controller in terms of robustness to inertia variations. This paper’s design method distinguishes between an adaptive linear neuron (ADALINE) and an RBFNN with a hidden layer. The linear integrator in a traditional PI controller can be thought of as an ADALINE, whereas the nonlinear integrator can be activated by using a hidden layer. Various experiments on the dSPACE MicroLabBox-based test bench are conducted to verify the effectiveness of the proposed method.

Repair based Constraint Handling Techniques for Microgrid Sizing and Energy Management Optimisation

ABSTRACT

Microgrid sizing and energy management system (EMS) optimisation problems have conflicting objectives while subjected to complex constraints. These problems are usually solved by using meta-heuristic algorithms, which are originally developed to solve unconstrained problems. Therefore, appropriate constraint handling technique (CHT) must be employed to solve constrained problems. It appears that use of CHTs in these problems is rare. This study proposes using two types of repair-based penalty approaches to solve a microgrid sizing and EMS problem. Cuckoo search algorithm is employed to solve the multi-objective optimisation problem, which minimises the levilised cost of electricity (LCOE) and dump load, while maximising the reliability of power supply. A case study based on the Westray Island standalone microgrid in Scotland is conducted to compare the effectiveness of the repair approaches, in terms of the objective function values and convergence speed.

Keywords: Constraint handling, Energy management system, Microgrid, Renewables, Repair methods, Sizing

Impact of Wave Energy Integration on Sizing and Energy Management of a Microgrid: Case study

ABSTRACT

This paper investigates the impact of integrating a wave energy converter (WEC) on sizing the battery and energy management of a microgrid utilising wind, solar and diesel generator at the generation side. A sequential co-optimisation model for sizing and energy management is proposed to minimise the levilised cost of electricity (LCOE) and dump load, while maximising the reliability of power supply. Cuckoo search algorithm is employed to solve the multi-objective optimisation problem. Moreover, a repair-based penalty approach is integrated for effective constraint handling. The Westray Island standalone microgrid in Scotland is considered as a case study. Annual hourly weather data of Westray Island and the demand profile are used to simulate the system in MATLAB environment. The numerical results show that the battery capacity is reduced when WEC is integrated in the microgrid. However, with the WEC integration, the LECO slightly increases with increased reliability of the power supply.

Keywords: Constraint handling, Energy management system, Microgrid, Sizing, Wave energy

test

asdsadasdas

New Concept for Current-Impressed Wireless Power Transfer to Multiple Independent Stainless-Steel-Enclosed Linear Actuator Tool Carriages

Linear actuators in, e.g., the food-processing or pharmaceutical industries are enclosed in stainless steel (SS) to facilitate thorough cleaning and disinfection.
Efficient wireless power transfer (WPT) through these SS enclosures realizes wireless electrification tool carriages and eliminates bulky, hard-to-clean power cables and cable carrier assemblies. This digest proposes a novel concept for current-impressed WPT through SS to multiple independent loads. The proposed method is thoroughly analyzed, optimized, verified in simulations, and compared against conventional voltage-impressed and current-impressed approaches, showing a clear complexity reduction and significant efficiency improvements, respectively. The final paper will contain a detailed experimental characterization of the proposed concept implemented in an industry-like, SS-enclosed linear actuator demonstrator system with two independent 100 W loads.

Evaluation of Drive Cycle Clustering Methods for Estimating the Drive Cycle Losses for the Design of Electrically Excited Synchronous Machines as EV Traction Drive

Evaluates different clustering methods for estimating the drive cycle losses of electric traction machines in a computionally efficient manner. Two motor types (electrically and permanent magnet excited synchronous machine, EESM and PMSM) are compared and it is highlighted for the first time in the literature that the EESM needs finer clustering in order to meet the same precision as the PMSM.

Torque Ripple Suppression of PMSM Speed Regualtion System Using Neural Network

In this paper, a neural network-based torque compensator is combined with a third order ESO in ADRC system to suppress the torque ripple. The third order ESO can effectively reject the low frequency disturbance, and the NN can output a feedforward compensation to optimize the torque and smooth the speed. Moreover, this NN compensator has a simple structure, which significantly reduces training cost time.

A Nonlinear PI Controller for Speed Control of Electric Drives Using Radial-Basis Function Neural Network

In this paper, a nonlinear proportional and integral (PI) controller based on Radial Basis Function Neural Network (RBFNN) is proposed for the speed control of electric drives. The Lyapunov function is employed in the design process to ensure system stability. The proposed nonlinear PI controller has a fixed proportional gain but a variable integral gain, which makes it outperforms than the conventional linear PI controller in terms of robustness to inertia variations. This paper’s design method distinguishes between an adaptive linear neuron (ADALINE) and an RBFNN with a hidden layer. The linear integrator in a traditional PI controller can be thought of as an ADALINE, whereas the nonlinear integrator can be activated by using a hidden layer. Various experiments on the dSPACE MicroLabBox-based test bench are conducted to verify the effectiveness of the proposed method.