Control Performance Assessment Subject to Multi-Objective User-Specified Performance Characteristics

User-specified benchmark has been used for control loop performance assessment. Structured closed-loop responses which are subject to various performance characteristics constraints are proposed in this work and served as a user-specified benchmarks against which the existing controller performance is compared. The performance characteristics constraints can be a mixture of variance upper bound, pole location, peak amplitude and model approximation constraint. An linear matrix inequality-based approach is used to formulate this multi-objective problem and a cone complementarity linearization algorithm is applied to find an optimal solution. Simulation studies and experimental results are provided to demonstrate application of the proposed benchmarks.