1 Introduction
In industrial production, due to the good motion and control characteristics of DC speed control system, DC speed control system is still widely used in many industrial sectors in my country that require high-performance controllable electric drives. From the perspective of control, DC speed control system is the foundation of AC speed control system. Strengthening the research on DC speed control system is conducive to promoting the further improvement of speed control system [1]. However, due to the influence of various factors (such as disturbances), DC speed control system may have serious instability problems during operation. At this time, the traditional control theory (such as PID control) cannot meet the requirements of high precision and high dynamic performance control. The current research on DC speed control system in my country mainly includes: comprehensive optimal control, compensated PID control, PID algorithm optimization, and some only use fuzzy control technology. Therefore, it can be seen that the main research point of DC speed control system is the application of intelligent control method. In recent years, neural network control has developed rapidly as an important part of intelligent control. As the most basic unit of neural network - single neuron, it has become the most basic control component in neural network control. This paper proposes a control strategy for DC speed regulation based on a combination of tracking differentiator, neural network and PID control to address the factors causing instability in DC speed regulation systems. Using Simulink simulation software, based on the S-function, the strategy leverages the filtering property of the tracking differentiator and the self-learning and adaptive properties of a single neuron to achieve fast, stable and real-time control of a dual-closed-loop DC speed regulation system with random disturbances through online learning control.
2. Problem Description
Statistics show that nearly 90% of loops in modern industrial process control still utilize traditional PID control strategies. However, a key issue in PID control is the tuning of the PID parameters. In practical applications, many controlled processes are relatively complex in principle.
The process parameters and even the model structure of DC speed control systems are complex and susceptible to changes in time and the working environment, making PID parameter tuning extremely difficult. For example, in production processes, DC speed control systems are not only complex in structure and parameters, but also present numerous challenges.
Disturbances, both in type and number, are not unique, leading to complex controller design, difficulty in suppressing disturbances, and poor system stability. The state-space description of this type of DC speed control system is as follows:
Disturbances. For this type of system, the design requirement for the controller is to first ensure system stability—that is, eliminate system disturbances—and then meet other system performance requirements. When W1 (t) and W2 (t) are deterministic disturbances, traditional methods can be used to design a corresponding disturbance rejection controller to ensure stable system operation. However, if W1 (t) and W2 (t) are both random disturbances, i.e., signals without obvious propagation patterns, it may cause uncertainty in the system model or parameters, complicating the system controller design. If traditional methods are still used to design the controller, it will be difficult to design and implement.
For DC speed control systems, the main disturbances include grid disturbances and load disturbances. Grid disturbances act within the current loop and can be suppressed promptly through current feedback, so their impact on speed is relatively small and can be disregarded. Load disturbances act outside the current loop but within the speed loop, and can only be suppressed by the speed regulator; therefore, the speed loop controller should be designed with good anti-interference capabilities. Furthermore, industrial control environments are complex and often contain various forms of interference sources, making external interference difficult to avoid. This paper, considering both load disturbances and external disturbances, designs a single-neuron PID controller and a tracking differentiator control system based on the S-function.
3 Controller Design
3.1 Single-Neuron PID Controller
Based on our hypothesis that human brain cells are adaptive, a complete single-neuron PID controller structure diagram is shown in Figure 1 below:
Figure 1. Structure diagram of a single-neuron PID controller
This paper focuses on a dual-closed-loop DC speed control system. From a control perspective, it applies the feedback principle and combines the unsupervised Hebb learning method of neural networks with the supervised Widrow-Hoff rule to derive a method for calculating weight values.
3.2 Tracking Differentiator (TD)
The tracking differentiator is a method for extracting differential signals proposed by Han Jingqing. It has good filtering performance, can arrange transition processes, and has phase lead functions. The tracking differentiator was initially proposed to solve the problem of reasonably extracting continuous signals and differentials from discontinuous or noisy measurement signals, and has gradually developed into a tracking differentiator that is easy to calculate.
This paper utilizes TD as the parameter input to arrange the transient process, resulting in a smooth input signal. In traditional PID controllers, the contradiction between speed and overshoot stems from directly applying the given input to the controller without any processing. The tracking differentiator can quickly track the input signal without overshoot, thus avoiding drastic changes in the control quantity and output overshoot caused by external disturbances in the input signal.
3.3 DC Speed Control System Using a Tracking Differentiator and a Single-Neuron PID Controller
The structure diagram of the dual closed-loop DC speed control system employing a tracking differentiator and a single-neuron PID controller is shown below:
Figure 2. Structure diagram of a dual closed-loop DC speed control system using a tracking differentiator and a single-neuron PID controller.
A dual-closed-loop DC speed control system requires the design of speed regulators and current regulators. As shown in the diagram, the inner loop is the current loop, and the outer loop is the speed loop. Considering that the outer loop—the speed loop—is the fundamental factor determining the control system, while the inner loop—the current loop—primarily modifies the operating characteristics of the controlled object to facilitate the control effect of the outer loop, the dual-closed-loop DC speed control system employs single-neuron PID control for the outer loop and retains traditional PI control for the inner loop, thus optimizing the control system.
4. Simulation Study
The parameters of the DC speed control system in this article are as follows: 220V, 136A.
Figure 3 Simulation model of dual closed-loop DC speed control system
Figure 4. Simulation curves of the system with only internal disturbances (1-Traditional PID; 2-Tracking differentiator combined with single neuron PID).
Figure 5. Simulation curves of the system with internal and external disturbances (1-Traditional PID; 2-Tracking differentiator combined with single neuron PID).
Based on the simulation results above, we can obtain the system performance indicators shown in the table below: The method for calculating the repetition value is as follows:
Table 1 System Performance Indicators
Control strategy System disturbances and parameters | Traditional PID control | Control strategy combining tracking differentiator and single-neuron PID control | |
Only exist Internal disturbances | Overshoot | 20% | 0 |
static error rate | 0.2 | 0 | |
Adjusting time | 1s | 0.5s | |
There are internal and external disturbances. | Overshoot | The system experienced oscillations. Unstable | Parameters are the same as above |
static error rate | |||
Adjusting time | |||
Analysis of the above data shows that under the control of the single-neuron PID controller and the tracking differentiator, the DC speed control system not only meets the speed requirements but also operates stably with zero overshoot and zero steady-state error. In contrast, traditional PID control results in system oscillations and a 20% overshoot, and the system response is slow. Therefore, the design of the single-neuron PID controller and tracking differentiator is not only simple and convenient, requiring no model of the controlled object, but this control strategy combining the tracking differentiator and single-neuron PID control is also feasible for dual-closed-loop DC speed control systems with multiple random disturbances.
5. Conclusion
The control strategy for a dual-closed-loop DC speed control system based on a single-neuron PID controller and a tracking differentiator is a linearized regulation process. It eliminates the need for modeling the controlled object during design, resulting in a simple structure, low computational complexity, and ease of implementation. Furthermore, it enables the system to maintain a good, fast, and stable response even under random disturbances. Compared to traditional PID controllers, the single-neuron PID controller is essentially a variable-coefficient proportional-integral-derivative composite controller with strong self-learning, adaptability, and robustness. The tracking differentiator is a controller that effectively addresses random disturbances and significantly improves the system's anti-interference capabilities.
6 References
[1] AC/DC Speed Control System and Matlab Simulation [M] Zhou Yuanshen, Song Yongying, China Electric Power Press, 2007.12.01 78-82
[2] Computer Control Systems [M] Kang Bo, Li Yunxia, Electronic Industry Press, 2011, pp. 1276-277
[3] Novel PID Control and Its Applications, [M] Tao Yonghua (ed.), Machinery Industry Press, 2nd edition, 2002, pp. 160-170.
[4] Advanced PID Control Simulation in MATLAB [M] Liu Jinkun, Electronic Industry Press, 2004, pp. 9256-270
[5] Wang Xiaodong, Chen Boshi, et al. Research on DC speed control system based on single neuron adaptive PID controller [J]. Electrical Drive, 1996, No. 4
[6] Han Jingqing. Active disturbance rejection controller and its application [J]. Control and Decision, 1998, 13(1): 19-23
