EM Servo Control for Axial Force Balance of Scroll Compressors Without Online Force Sensors

Authors:
Publisher website
Google Scholar
Abstract
When the electromagnetic (EM) force ${F}_{\mathrm{M}}$ is used to balance the dynamical axial gas separation force ${F}_{\text{gas}}$ of scroll compressors, the control effect, system complexity, reliability, and cost are severely restricted by the dynamic force sensor systems. For this consideration, a scroll compressor without online force sensors is designed, and an innovative EM force control method based on “EM force-current ( ${F}_{\mathrm{M}}{\rm{ - }}{i}_{\mathrm{M}}$ )” nonlinear model is proposed. The “ ${F}_{\mathrm{M}}{\rm{ - }}{i}_{\mathrm{M}}$ ” nonlinear relationship changes with frequency and EM gap, so it is difficult to determine the relationship online. Even when the target EM force is precise, it is difficult to obtain the corresponding control current to generate the actual EM force. Given this difficulty and challenge, this article creatively puts forward a method. When the target EM force is always kept as the final target curve, the “ ${F}_{\mathrm{M}}{\rm{ - }}{i}_{\mathrm{M}}$ ” relationship, used to decode the control current, is adjusted by adjusting the estimated value of the EM gap to control the actual EM force. The best dynamic contact seal and high-quality steady-state operation can be realized by optimizing the estimated EM gap value. This method consists of two stages: set up an offline database and carry out online control. At the offline stage, the “ ${F}_{\mathrm{M}}{\rm{ - }}{i}_{\mathrm{M}}$ ” nonlinear relations with different frequencies and EM gaps are sampled with force sensors and then stored using Prandtl–Ishlinskii (PI) model. At the online stage, no online feedback force sensors are needed. The current needed to produce exact F M is decoded from the chosen “ ${F}_{\mathrm{M}}{\rm{ - }}{i}_{\mathrm{M}}$ ” PI model established at the offline stage. The creative “ ${F}_{\mathrm{M}}{\rm{ - }}{i}_{\mathrm{M}}$ ” model choosing method is proposed, which relies on the iterative estimations of EM gap value. The results of the quasi-prototype experiment show that the aforesaid control method is feasible.
Funding Information
  • Natural Science Foundation of Zhejiang Province, China (LY17E070002, Y21E070002)
  • Public Projects of Zhejiang Province, China (LGG22E070006)

This publication has 24 references indexed in Scilit: