The excitation system is an important component of the synchronous generator , and its characteristics directly affect the reliability and stability of the synchronous generator system. At present, many small and medium-sized generator sets in China, especially those in remote rural areas, still use old-fashioned excitation control systems produced in the 1970s and 1980s. These systems have large sparks, are prone to wear, require a lot of maintenance, and have slow response speed, which can no longer meet the requirements of the increasingly developed power system. With the widespread application of high-power semiconductor devices and the rapid development of computer technology, it is imperative to replace the old electromechanical excitation control system with a new type of excitation control system developed with modern power electronics technology. Over the years, it has been recognized that as a generator excitation system, hardware and control strategy are two important interconnected parts [1].
This paper designs a parameter self-tuning fuzzy PID control algorithm for the nonlinear system control of excitation control system using parameter tuning methods based on intelligent control theory, enabling it to meet the operational performance requirements of synchronous generators. Generator parameters are sampled via AC sampling followed by FFT calculations to obtain parameters such as generator voltage, current, power factor, active power, reactive power, and harmonic voltage. The generator excitation voltage is controlled using power MOSFETs and PWM technology, thus achieving voltage control and regulation. The hardware structure uses TI's TMS320F2812 chip as the control core, and parameter setting and data exchange are performed through various methods including human-machine interface, DSP via serial communication, and network communication. A DSP-based software program is designed and debugged using CCS 2.21, and finally, field tests are conducted with a synchronous generator, with recorded experimental data analyzed and compared. Experimental results show that compared with traditional PID control, this regulator has better dynamic and static performance, faster adjustment speed, and a smoother adjustment curve; its main performance indicators meet or exceed national standards.
Hardware of the excitation control system; Overall hardware structure of the excitation system
The excitation system of the generator mainly consists of the main excitation system and the excitation controller. The excitation controller in this paper is designed for the main excitation system to be third harmonic excitation. The main excitation system is characterized by the power supply being taken from the terminal voltage of the third harmonic winding of the generator, which is rectified by the rectifier circuit and applied to the excitation winding as DC power. Compared with other excitation systems with rotating motors, this excitation method eliminates the need for a rotating exciter. Therefore, this excitation method is gradually being widely adopted due to its simple structure, low cost, and reliable performance [2]. The overall hardware structure diagram of the system is shown in Figure 1.
Figure 1 System overall hardware structure diagram
Control circuit section of the excitation system
Main control chip [3]
The core of this controller is the TMS320F2812, a high-speed, high-precision industrial control DSP chip launched by TI in recent years. It features fast processing speed, a clock frequency of up to 150MHz, an instruction cycle of less than 6.67ns, and low power consumption (core voltage 1.8V, I/O port voltage 3.3V). Utilizing a Harvard bus architecture, it boasts powerful operational capabilities, rapid interrupt response and processing, and a unified register programming mode, enabling 16×16-bit and 32×32-bit multiply-accumulate operations, as well as two 16×16-bit multiply-accumulate operations. On-chip memory includes 8KB×16-bit flash memory, 1KB×16-bit OTP read-only memory, two 4KB×16-bit single-port random access memory (RAM), one 8KB×16-bit RAM, and two 1KB×16-bit RAM. The external memory interface allows for expansion up to 1MB of memory. The peripheral expansion module (PIE) supports 96 peripheral interrupts, with 45 available. Two enhanced event manager modules (EVA and EVB) provide a complete set of functions and features for motion control and motor control applications. Each event manager module includes a general-purpose timer (GP), a compare unit, a capture unit, and an orthogonal encoder pulse circuit. Peripherals include three 32-bit CPU timers, a 16-channel 12-bit ADC (single conversion time of 200ns, single-channel conversion time of 60ns), and features not only a Serial Peripheral Interface (SPI) and two UART interface modules (SCI), but also an improved local area network (ECAN), a multi-channel buffered serial port (MCBSP), and a serial peripheral interface mode.
Input detection module [4]
Its function is to acquire voltage and current signals. Voltage is converted into current information through a current-type precision transformer, and current is converted into current information through a current transformer. Then, a signal conditioning circuit composed of operational amplifiers generates a 0-3V voltage signal, which is sent to the AD conversion channel of the DSP to convert the analog quantity into a digital quantity, which is then placed into a designated unit for program use.
PWM wave IGBT drive circuit
The driver circuit of an IGBT module plays a decisive role in its operating efficiency. The requirements for the driver circuit are to minimize switching losses while also providing a good drive waveform with low oscillation and minimal overshoot. When the IGBT operates in switching mode, its gate must be charged and discharged, and driving the IGBT requires a certain amount of charge; that is, the driver circuit must provide a certain driving voltage and driving power.
Excitation control module
The function is to control the excitation current of the exciter using a controllable PWM wave generated by a DSP. Diesel generators are small generators, with excitation currents typically around ten amperes. Therefore, this paper uses a low-power MOSFET (Metal-Oxide-Semiconductor Field-Effect Transistor) as the switching element. The voltage regulator's operating voltage comes from the generator's terminal voltage. The voltage regulator can only operate when there is voltage across the generator terminals, requiring a separate excitation circuit in the main circuit to provide the excitation current. Once the generator output voltage reaches a certain value, it automatically switches to the MOSFET-controlled circuit. Therefore, this paper adds an intermediate relay to the main circuit. Its normally closed contact is connected in parallel across the drain and source terminals of the MOSFET, short-circuiting the MOSFET and providing a closed excitation circuit for excitation. Before the relay operates, the rectified third harmonic voltage is entirely output to the starting resistor and the excitation winding of the exciter. As the generator output voltage rises rapidly, the intermediate relay coil is energized and begins to operate. After the relay is activated, its normally closed contact automatically opens, the MOSFET starts to conduct, and the excitation circuit automatically switches from an uncontrollable system to a closed-loop system controlled by the DSP.
Human-computer dialogue module
Its function is to enable human-machine interaction and parameter modification through a keyboard and LED display circuit or a host computer monitoring method. The software structure design and implementation of the excitation regulator.
Overall System Software Flow
Generator excitation control is a fast, real-time closed-loop regulation that requires a high response speed to changes in generator terminal voltage to maintain the voltage at a given level. Simultaneously, to ensure safe generator operation, the excitation controller must also possess limiting functions that protect the generator and excitation system, such as forced excitation limiting and under-excitation limiting. The regulation, limiting, and control functions of the microprocessor-based excitation controller are all implemented through software.
In the software design of the excitation regulator, there are three main design ideas: first, the software must meet the needs of real-time control; second, the software should give full play to the instruction and hardware features of the TMS320F2812; and third, the software should have strong flexibility, versatility and reliability, and take into account the needs of future functional expansion. The main program flowchart is shown in Figure 2 [5,6].
Figure 2 Main Program Flowchart
PWM waveform generation module
The PWM module is designed to minimize CPU overhead and user intervention when generating pulse width modulation waveforms for motor and motion control applications. In this system, only one PWM output is needed for motor control. However, for system reliability, two PWM outputs are used. This allows for the simultaneous generation of two PWM waveforms with opposite duty cycles, and also enables the PWM output to be restored without major hardware modifications if one output fails. This system configures the IOPA6/PWM1 and IOPA7/PWM2 pins of the TMS320F2812 in basic mode, i.e., configured for PWM1 and PWM2 output functions. PWM generation is accomplished by the full comparator unit under the control of the corresponding control registers.
Figure 3. Structure diagram of fuzzy parameter self-tuning PID excitation control system
Parameter self-tuning fuzzy PID controller [7]
The adoption of DSP solved the speed problem of the excitation regulator system. However, in addition to speed requirements, the system also requires control accuracy. To achieve good control accuracy, a good control algorithm must be adopted. A parameter self-tuning fuzzy PID control algorithm is used in the microcomputer excitation regulator. Because the computer output increment has a small impact from erroneous actions and is easy to achieve automatic, disturbance-free switching, it utilizes the idea of fuzzy inference to perform online self-tuning of the three parameters kp, ki, and kd of the PID controller based on different e (deviation amount) and ec (deviation change amount). This forms the parameter self-tuning fuzzy PID controller. Its structure is shown in Figure 3. It can automatically adjust the three parameters of the PID controller online in real-time control, enabling the parameter self-tuning fuzzy PID controller to operate in an optimal state.
Considering the actual working conditions of the synchronous generator excitation regulator, the sampling period t was set to 20ms. A terminal voltage control experiment was conducted on the synchronous generator in the laboratory. The measured values were: kp=32.5, ki=2.1, kd=24.5. The synchronous generator terminal voltage exhibited good excitation boost characteristics. Furthermore, when kpmin=8.2, kpmax=63.6, kimin=1.1, kimax=6.1, kimin=1.1, kimax=6.1, kdmin=3.8, and kdmax=482, the terminal voltage also showed a good excitation response curve. Based on the measured PID parameters, the fuzzy controller design method, and the fuzzy parameter self-tuning rules, a parameter self-tuning fuzzy PID excitation controller can be designed.
Experimental Results and Analysis
Comparison of terminal voltage waveforms of traditional PID-controlled synchronous generators and parameter-self-tuned fuzzy PID-controlled synchronous generators: In the experimental results graph of this system, the horizontal axis numbers represent time, approximately 250 represents 1 second, and the vertical axis represents the terminal voltage of the synchronous generator, in V. The frequency of the synchronous generator is 50 Hz.
Figure 4. Voltage waveforms of PID control with large step-down and fuzzy PID control. Figure 5. Voltage waveforms of PID control with boost and fuzzy PID control.
Two typical experiments, one involving significant voltage reduction and the other a voltage boost, were selected to compare and analyze the control effects of traditional PID control and parameter self-tuning fuzzy PID control. The experimental waveforms in Figures 4 and 5 show that the designed synchronous generator excitation controller exhibits good steady-state performance. The results demonstrate that, compared to the traditional PID controller, the fuzzy parameter self-tuning PID controller has a faster response speed, smaller overshoot, smoother transition process, and better system tracking performance.
Conclusion
This paper designs an excitation controller for a synchronous generator based on the TMS320F2812 chip. It employs a parameter self-tuning fuzzy PID control strategy and uses MOSFET power devices for PWM regulation of the excitation current. The controller fully utilizes the DSP's fast computing power and abundant on-chip peripheral resources, integrating the DSP with other functional modules into a cohesive whole. An ECAN fieldbus is used to construct the network, ensuring high reliability and stability of communication. Field experimental results show that this controller has better dynamic and static performance than traditional PID controllers, with faster adjustment speed and a smoother adjustment curve. Its main performance indicators meet or exceed national standards.
