Abstract: This paper introduces an intelligent fuzzy control system based on the 8098 microcontroller and using a heating furnace as the controlled object. The rule-based self-optimization fuzzy control algorithm is studied in depth. Simulation results show that the system has good control effect, high steady-state accuracy, and small overshoot. Keywords: Heating furnace; Fuzzy control; Microcontroller; Simulation 1 Introduction Heating furnaces are widely used in metallurgical, chemical, and other industrial production processes. The temperature of the heating furnace is an important indicator of the production process, and the quality of temperature control directly affects the product quality. Heat treatment heating furnaces are a process that can improve the properties of metal materials and their products (such as machine parts, tools, etc.). Depending on different requirements, materials and their products are heated to a suitable temperature and held at that temperature, then cooled using different methods to change their internal structure to obtain the required properties. Heat treatment is one of the important means to improve the quality of metal materials and their products. Heat treatment heating furnaces have nonlinear and time-varying characteristics such as large inertia and pure time delay. The opening and closing of the furnace door, the heating material, the ambient temperature, and the power grid voltage all affect the control process. Conventional control based on precise mathematical models, such as PID control, is difficult to guarantee the heating process curve requirements. As a major branch of nonlinear control, fuzzy control can be well applied in the above temperature control system. Fuzzy control is one of the branches of intelligent control. It has the following characteristics: (1) It is a nonlinear control method with a wide working range and wide applicability. It is particularly suitable for the control of nonlinear systems. (2) It does not depend on the mathematical model of the object. For complex objects that cannot be modeled or are difficult to model, human experience and knowledge can be used to design a fuzzy controller to complete the control task. Traditional control methods require the mathematical model of the controlled object to be known before the controller can be designed. (3) It has an inherent parallel processing mechanism, exhibits strong robustness, and is not sensitive to changes in the characteristics of the controlled object. The design parameters of the fuzzy controller are easy to select and adjust; the algorithm is simple, fast, and easy to implement, and does not require much control theory knowledge. 2 Design of Fuzzy Controller This control system mainly completes the functions of data acquisition, temperature display, furnace temperature control, fault detection, and alarm. The intelligent fuzzy controller is completed by a microcontroller and uses a rule-based self-optimization control algorithm for process control. The heating furnace adopts bidirectional thyristor control. The microcontroller outputs the on/off rate control signal to generate the zero-crossing trigger pulse of the thyristor. The core of the entire system is a fuzzy controller, with the 8098 microcontroller as its main component. It, along with several expansion circuits (program memory, data memory, address latches, address decoders, etc.), constitutes the processor module. The mV signal output from the thermocouple is converted into a standard 0-10V signal by the transmitter chip, then converted by an A/D converter and sent to the microcontroller. Based on various input commands, the microcontroller calculates the control quantity using a fuzzy control algorithm, outputs a pulse trigger signal, and drives a bidirectional thyristor through a zero-crossing trigger circuit, thereby controlling the heat treatment furnace. In addition, the intelligent controller also includes a hardware watchdog circuit, a fault detection circuit, a digital display circuit, and a power supply. The hardware block diagram of the intelligent fuzzy controller is shown in Figure 1. The main program of the fuzzy controller includes initialization, keyboard management, and calls to the control and display modules. The implementation of functions such as temperature signal acquisition, digital filtering, scaling transformation, control algorithms, and temperature display is completed by various subroutines. The main software flow is as follows: The sampling period is generated using the 8098 microcontroller's timer T0 and software counter. When the period expires, the program switches to the control module, calling the A/D conversion, digital filtering, and scaling modules to obtain the furnace temperature feedback signal. The control quantity is calculated based on the deviation and the rate of change of the deviation, and a pulse signal is output to control the zero-crossing trigger. Start, stop, and setpoint are generated via external interrupts using the keyboard; when a key is pressed, the interrupt service routine is called. The program flow is shown in Figure 2. 3. Research on Fuzzy Control Algorithm The heat treatment furnace in this system is a large inertial system with pure time delay. Conventional control based on a precise mathematical model is insufficient to guarantee the heating process curve requirements. Therefore, a rule-based self-optimization algorithm from the fuzzy control algorithm is selected. The basic principle of the algorithm uses control rules described by analytical expressions, which is simple, convenient, and easy to process. The two-dimensional control rule-based self-optimization algorithm can be summarized by the analytical expression: where E, C, and U are fuzzy variables after quantization and fuzzification, and their corresponding domains of discourse are error, rate of change of error, and control quantity, respectively; a is the adjustment factor. As can be seen from the control rule described by equation (1), the control action depends on the error and the rate of change of the error. By adjusting the value of 'a', the different weightings of the error and the rate of change of the error can be changed. Once the value of 'a' is determined, it will not change during the entire control process. However, in actual systems, the weightings of the error E and the error C in the control rule have different requirements under different states. For example, when the error is large, the main task of the control system is to eliminate the error, so the weighting of the error should be larger. When the error is small, the main task of the control system is to stabilize the system as soon as possible and reduce overshoot, so the weighting of the rate of change of the error in the control rule should be larger. In order to obtain good control performance, the value of 'a' must be adjustable during the control process, that is, the control rule can be corrected online during the control process. Equation (2) uses the optimization design method to correct 'a' online. The control principle of the system is shown in Figure 3. 4 Performance Analysis After the simulation operation, the system works stably and is easy to operate. All indicators have met the design requirements. Figure 4 shows the temperature curve of a heating furnace. It can be seen from the figure that the curve has good tracking performance, high steady-state accuracy, and small overshoot. 5 Conclusion This paper systematically introduces the temperature control system of a heat treatment heating furnace implemented with a microcontroller. Since the controlled object is a large inertial element with pure time delay, the use of an intelligent fuzzy controller can achieve the ideal control effect. The experimental results show that the system has the following characteristics: (1) The control scheme is reasonable, the steady-state accuracy is high, and the overshoot is small. (2) The structure is simple, the debugging is convenient, the anti-interference is strong, and the robustness is good.