Topology Optimization for Harmonic Excitation Structures with Minimum Length Scale Control Using the Discrete Variable Method

Continuumtopology optimization considering the vibration response is of great value in the engineering structure design.The aimof this studyis toaddress the topological designoptimizationof harmonic excitationstructureswith minimumlength scale control to facilitate structuralmanufacturing.Astructural topology design based on discrete variables is proposed to avoid localized vibration modes,gray regions and fuzzy boundaries in harmonic excitation topology optimization.The topological design model and sensitivity formulation are derived.The requirement of minimum size control is transformed into a geometric constraint using the discrete variables.Consequently,thin bars,small holes,and sharp corners,which are not conducive to the manufacturing process,can be eliminated from the design results.The present optimization design can efficiently achieve a 0–1 topology configuration with a significantly improved resonance frequency in a wide range of excitation frequencies.Additionally,the optimal solution for harmonic excitation topology optimization is not necessarily symmetric when the load and support are symmetric,which is a distinct difference fromthe static optimization design.Hence,one-half of the design domain cannot be selected according to the load and support symmetry.Numerical examples are presented to demonstrate the effectiveness of the discrete variable design for excitation frequency topology optimization,and to improve the design manufacturability.

Simplified Deep Reinforcement Learning Based Volt-var Control of Topologically Variable Power System

The high penetration and uncertainty of distributed energies force the upgrade of volt-var control(VVC) to smooth the voltage and var fluctuations faster. Traditional mathematical or heuristic algorithms are increasingly incompetent for this task because of the slow online calculation speed. Deep reinforcement learning(DRL) has recently been recognized as an effective alternative as it transfers the computational pressure to the off-line training and the online calculation timescale reaches milliseconds. However, its slow offline training speed still limits its application to VVC. To overcome this issue, this paper proposes a simplified DRL method that simplifies and improves the training operations in DRL, avoiding invalid explorations and slow reward calculation speed. Given the problem that the DRL network parameters of original topology are not applicable to the other new topologies, side-tuning transfer learning(TL) is introduced to reduce the number of parameters needed to be updated in the TL process. Test results based on IEEE 30-bus and 118-bus systems prove the correctness and rapidity of the proposed method, as well as their strong applicability for large-scale control variables.