Design of high-power constant-current discharge system for batteries

    Storage batteries have been widely used in various industries such as military, railways, communications, and electricity, and have gradually become a very important backup power source in daily life [1-2]. In the production and use of batteries, performance testing is an indispensable task. Constant current discharge is the most direct and effective method to study battery performance. At present, there are two methods for discharging batteries: using fixed resistors, variable resistors, resistor boxes, etc. as discharge loads, which requires manual adjustment of the discharge current and has low control accuracy [3]; using a switching power supply boost circuit method to control the voltage applied to the load by adjusting the duty cycle. This method has large switching losses and current pulsation [4].
    In order to solve the above problems, a battery performance comprehensive tester with power MOSFET tubes as electronic loads was developed, which can automatically control and monitor the entire discharge process, record the changes in voltage and current during the discharge process with sufficient density, and display them in a graphical interface.
1 Overall design
    The high-power constant current discharge system of batteries mainly includes a host computer, a serial port unit, a control unit, a drive unit, a discharge unit, and a data acquisition unit. Figure 1 is a system structure block diagram.

2 Control unit design
    The traditional PID controller has a simple structure and high precision. It has been widely used in the field of industrial process control and has achieved good control effects. However, for time-varying and nonlinear systems, even if a set of satisfactory PID parameters is adjusted for the controlled object, it is difficult to ensure good control performance when the object characteristics change [5]. After repeated experiments on the constant current discharge control system, it is concluded that the fuzzy PID control method is more appropriate.
    The fuzzy PID controller is mainly composed of two parts: the PID controller and the fuzzy controller. The input of the fuzzy controller is the error e and the error change rate ec. Then, the fuzzy reasoning method is used to adjust the PID parameters Kp, Ti, and Td online to meet the different requirements of the error e and the error change rate ec at different times for the controller parameters, so that the controlled object has good dynamic and static performance [6]. Its structure is shown in Figure 2.

    The control chip used in this system is the single-chip microcomputer C8051F020. The input of the controller is the voltage command sent by the host computer, and the output of the system is the output voltage of the drive unit. The actual voltage signal fed back is compared with the given voltage signal to obtain a voltage error signal. The fuzzy PID controller calculates the precise PID parameter adjustment factor to achieve the purpose of adjusting the PID controller parameters. Then the control signal is calculated, and the control signal is converted into the gate voltage of the MOSFET through the operational amplifier circuit of the drive unit to adjust the output current of the discharge unit.
3 Research and design of constant current discharge scheme
3.1 Selection of discharge type
    At present, the methods of discharging batteries are mainly divided into two categories: energy consumption type and energy feedback type. The energy consumption type is to transfer the DC power flowing into the electronic load to a special DC/AC converter (inverter circuit) and then send it back to the AC power grid. The main advantages of this type are saving energy, saving space, and not requiring cooling [7]. However, its discharge current is volatile, has low accuracy, and is prone to cause harmonic pollution to the power grid. The energy consumption type refers to consuming the electric energy discharged by the battery through the power tube and releasing it as heat or other forms of energy. The main advantages of the energy consumption type are simple structure, economical and practical, high precision, etc. The disadvantages are energy consumption and the need for a good heat dissipation system. However, for the battery performance test, since the accuracy of constant current discharge is guaranteed, the energy consumption type discharge method is used for design.
3.2 Research and analysis of different discharge schemes
    For energy consumption type constant current discharge, there are currently three schemes that can be used to achieve it.
    Scheme 1: Direct current sampling method, that is, the total current flowing into the electronic load is detected by the Hall current sensor, and then compared with the set current to determine whether the system setting value is reached. If it is not reached, it is necessary to adjust the given voltage Vg through a certain algorithm, and finally stabilize the load current at the set value. Scheme 2: Resistor sampling feedback method, that is, a sampling resistor is connected in series to the source of the MOSFET tube to convert the current into a voltage and feed it back to the inverting end of the high gain error amplifier . The simple schematic diagram of the resistor sampling feedback method is shown in Figure 3. Input a given voltage signal at the same-direction end. If the voltage on Rd is less than Vg, that is, the voltage at the reverse end of the operational amplifier is less than the voltage at the same-direction end, the output voltage of the operational amplifier increases, which increases the on-current of the MOSFET; if the voltage on Rd is greater than the given value Vg, the output of the operational amplifier decreases, which reduces the on-current of the MOSFET. In this way, the current is finally maintained at a constant given value, thus achieving constant current operation.

    Scheme 1 has a large current variation range and is suitable for large current discharge; its disadvantage is that the response speed is relatively slow. Scheme 2 has the advantages of fast response speed and simple structure; however, since the power of the sampling resistor is generally small, the current of the electronic load is greatly limited and easily affected by external interference, making it unsuitable for large current discharge.
    In view of the advantages and disadvantages of the above two schemes, Scheme 3 is proposed: connect several discharge branches of Scheme 2 in parallel, that is, adopt the design idea of ​​multiple discharge branches in parallel [8]. However, due to the difference in parameters of the power switch tube and the branch control circuit, the current flowing through each branch switch tube cannot be completely equal. Therefore, it is necessary to design a special current balancing circuit to achieve current balancing between branches, thereby achieving power balancing of branches and avoiding excessive power in some branches, which may burn out the power switch tube [9-10]. In this way, high-power discharge can be carried out while ensuring that each branch operates within a smaller current range. However, a current balancing circuit is required to control each branch, and the current balancing circuit is composed of many ideal devices such as resistors and operational amplifiers with the same parameters. Since the parameters of actual devices are not exactly the same, a large error is introduced into the current balancing circuit.
3.3 Final design of discharge scheme
    The current-sharing circuit will bring inevitable errors to the system. Therefore, the current-sharing circuit is no longer used. Instead, the single-chip microcomputer is used to directly control the given voltage Vg of each branch, so that the given voltage of each branch is equal to achieve current balancing of each branch, so as to eliminate the error caused by the current-sharing circuit. In order to further improve the accuracy of the system, the total current discharged by the battery is collected by the Hall current sensor and fed back to the C8051F020 single-chip microcomputer. The single-chip microcomputer adjusts the size of the given voltage Vg of each branch through the fuzzy PID control algorithm to control the current of each branch, so as to achieve the purpose of accurately controlling the total current.
4 Experimental results
    In order to more directly test the effect of constant current discharge, this experiment uses a voltage signal with a change of 11 V~13 V instead of the battery. The system uses 5 discharge branches to discharge in parallel, and the total current is set to 15 A. Figure 4 is the experimental results of constant current discharge.

    The battery constant current discharge system based on C8051F020 single chip microcomputer adopts the idea of ​​multi-branch parallel connection to achieve high-power discharge; the current size of each branch is controlled by software to ensure that each branch can work within the rated power, improving the stability of the system; the fuzzy PID control method is adopted to improve the accuracy of system discharge. In addition, the discharge process can be monitored and recorded in real time by the host computer, which provides convenience for the analysis of battery performance.
References
[1] Wang Chengshan, Wang Shouxiang. Research on several issues of distributed power generation and energy supply system [J]. Automation of Electric Power Systems, 2008, 32(20): 1-4.
[2] Fu Shudun. Suggestions for the development of smart grid
in China [J]. Automation of Electric Power Systems, 2009, 33(10): 23-26. [3] Chen Jie, Xu Jianhong. Failure mechanism and detection of valve-controlled sealed lead-acid batteries [J]. Power Supply Technology, 1999, 23(6): 332-334.
[4] Gao Jiaying, Gao Yufeng, Liu Yalong. Research on constant current discharge equipment based on new electronic load [J]. Power Supply Technology Application, 2007, 10(8): 48-51.
[5] Liu Jinkun. Advanced PID control and simulation by Matlab [M]. Beijing: Publishing House of Electronics Industry, 2003.
[6] Huang Zhengwu. Research on DC electronic load control algorithm [D]. Liuzhou: Guangxi Institute of Technology, 2011.
[7] Wang Hao. Design of electronic load for fuel cell test system [D]. Chengdu: University of Electronic Science and Technology of China, 2007.
[8] Yan Zhifeng, Ma Xiaojun, Wei Shuguang. Research on a novel constant current discharge device for batteries [J]. Journal of the Armored Forces Engineering Academy, 2006, 20(3): 90-92.
[9] Zhang Suwen. High Frequency Electronic Circuits (Second Edition) [M]. Beijing: Higher Education Press, 1984.
[10] HOFER P, KARRER N. Paralleling intelligent IGBT power modules with active gate-control led  current balancing [C]. 27th Annual IEEE Power E1ectronics Specialists Conference, 1995: 342-346.