Abstract : Directly driving induction motors using general-purpose frequency converters operating in constant voltage-frequency ratio mode is widely used in industrial production. Improving the control performance of the constant voltage-frequency ratio variable frequency speed control system without altering its original structure is of great significance. This paper, based on the characteristics of constant voltage-frequency ratio control, treats the general-purpose frequency converter and induction motor as a whole—the constant voltage-frequency ratio variable frequency speed control system—and presents the corresponding mathematical model. It also derives inverse system models applicable to both compensated and uncompensated constant voltage-frequency ratio operation modes. Furthermore, a neural network inverse system is constructed and connected in series with the constant voltage-frequency ratio variable frequency speed control system to form a pseudo-linear system. A linear closed-loop regulator is then designed to achieve high-performance control. Experimental results demonstrate that the neural network inverse control method can enable the constant voltage-frequency ratio variable frequency speed control system to achieve excellent operating performance. Keywords : Induction motor; Frequency converter; Neural network; Inverse control 1 Introduction Due to the small size, light weight, versatility, and high reliability of general-purpose frequency converters, variable frequency speed control systems composed of frequency converters and induction motors have been widely used in the past 20 years, and are showing a strong trend of replacing DC speed control systems, enjoying popularity in industries such as machinery, chemical engineering, metallurgy, and light industry. However, with the continuous improvement of production requirements, some new problems have arisen in such a general-purpose frequency converter + induction motor system (including a load, referred to as a variable frequency speed control system in this paper). On the one hand, many existing systems need to be improved to enhance their performance. Eliminating existing general-purpose frequency converters is impractical and detrimental to system upgrades. On the other hand, due to their low price and high reliability, newly constructed systems still largely utilize general-purpose frequency converters. Furthermore, when implementing coordinated control of multi-motor systems or network-distributed control of the entire production line, further control of such variable frequency speed control systems driven by general-purpose frequency converters is required (often to further improve the performance of the entire system or production line). Commonly used PID control methods cannot achieve satisfactory control results. Therefore, how to further improve the control performance of a variable frequency speed control system without changing its traditional structure (i.e., treating the frequency converter and induction motor as a whole) is an urgent problem to be solved. [b][align=center]For details, please click: Neural Network Inverse Control of Constant Voltage-Frequency Ratio Variable Frequency Speed ​​Control System[/align][/b]