A BMS system SOC calculation method based on metering chip

Abstract

The present invention relates to a method for calculating the SOC of a BMS system based on a metering chip. The metering chip is connected to an MCU and a voltage reference source chip, and the method includes the following steps: the MCU sends a start metering command to the metering chip, and the metering chip calculates the charge and discharge amount containing the voltage coefficient after receiving the start metering command; the MCU reads the charge and discharge amount containing the voltage coefficient calculated by the metering chip according to the acquisition timing defined by the BMS; a single cell power matrix is generated based on the charge and discharge amount containing the voltage coefficient; at the same time, a temperature sampling point distribution matrix is generated; the single cell power matrix and the temperature sampling point distribution matrix are expanded into matrices with the same rows and columns; the two expanded single cell power matrices and temperature sampling point distribution matrices are subjected to Coleman filtering in a one-to-one correspondence to obtain the single cell SOC value. The present invention uses a separate electric energy metering chip to be solely responsible for the integration work, which can greatly save the computing power of the MCU.

Description

Technical Field

[0001] The present invention relates to the field of new energy lithium battery energy storage technology, and in particular to a BMS system SOC calculation method based on a metering chip. Background Art

[0002] With China's policy shift towards the new energy industry, lithium batteries and BMS systems are being widely used in many fields. However, estimating the SOC of lithium batteries, a crucial function of BMS systems, is difficult to accurately calculate.

[0003] 1. The ampere-hour integration method commonly used today essentially accumulates the collected current signal at short, uniform time intervals. However, commonly used MCUs are limited by hardware conditions and find it difficult to sample current values at a high frequency, resulting in a rough integration effect and large errors. Common BMS systems typically use a current sampling frequency of 10 to 200 Hz. Excessively high sampling frequencies can occupy a significant amount of resources and affect other system tasks. If an external high-frequency ADC chip is used specifically for current sampling, the integration process still needs to be completed by the MCU. A higher sampling frequency also requires higher computing power, which also consumes a significant amount of resources.

[0004] 2. Because the MCU is simultaneously processing other tasks while calling the timer to sample the current signal, frequent interrupts in these other tasks can easily disrupt timing, causing inaccurate timing and significantly increasing SOC errors. Since integral calculations require extremely precise timing, setting a high interrupt priority will frequently interrupt other tasks. Furthermore, a high sampling frequency can disrupt normal tasks. A lower priority will likely result in frequent interruptions from other tasks, impacting accuracy. This suggests that the sampling frequency in the MCU is inherently limited by the system.

[0005] 3. Because the MCU load may vary at different times, the timer times are not the same at high and low loads. While the timer consistency may be very good over a short period of time, over a longer period of time, the timer will have slight time differences due to the varying MCU loads. This will manifest as a shift in the slope of the SOC curve, leading to significant deviations in the overall SOC value.

[0006] 4. Because the MCU is at the core of the entire circuit board and cannot be effectively isolated from other components, it is severely affected by heat and causes temperature drift. Temperature drift is sufficient to produce non-negligible errors for more stringent measurement. The best solution is to isolate the chip responsible for measurement from the main circuit to minimize its exposure to the main circuit's heat.

[0007] 5. In the process of counting SOC, the BMS system will also count the total charge and discharge amount of the system. This process is also completed by ampere-hour integration. Therefore, a large error in ampere-hour integration affects not only the accuracy of SOC, but also the statistics of the charge and discharge amount of the entire system.

[0008] 6. To accurately estimate the SOC of each battery, in addition to accurately calculating the ampere-hour integral, the battery temperature must also be accurately collected. The temperature effect must then be fitted into the time-integrated value as a coefficient. However, in reality, the temperature sampling points within the battery pack do not correspond 1:1 to each battery, and the collected temperature data cannot be mapped to each battery. Therefore, the battery SOC matrix and temperature matrix must be introduced and amplified to obtain two matrices with the same number of rows and columns. Kalman filtering is then performed on each corresponding data in the two matrices to obtain the accurate SOC value. Summary of the Invention

[0009] In view of the above-mentioned deficiencies in the prior art, the technical problem to be solved by the present invention is to provide a BMS system SOC calculation method based on a metering chip.

[0010] The technical solution adopted by the present invention to achieve the above-mentioned purpose is: a BMS system SOC calculation method based on a metering chip, wherein the metering chip is connected to an MCU and a voltage reference source chip, and the method comprises the following steps:

[0011] The MCU sends a start measurement command to the measurement chip, and the measurement chip calculates the charge and discharge amount including the voltage coefficient after receiving the start measurement command;

[0012] The MCU reads the charge and discharge amount including the voltage coefficient calculated by the metering chip according to the acquisition timing defined by the BMS;

[0013] Generate a single cell power matrix based on the charge and discharge quantities containing the voltage coefficient; and at the same time, generate a temperature sampling point distribution matrix;

[0014] Expanding the single cell power matrix and the temperature sampling point distribution matrix into matrices with identical rows and columns;

[0015] The two amplified single cell battery power matrices and temperature sampling point distribution matrices are subjected to Coleman filtering in a one-to-one correspondence to obtain the single cell battery SOC value.

[0016] The calculation of the charge and discharge capacity including the voltage coefficient comprises the following steps:

[0017] The metering chip collects the current value of the total positive circuit of the entire battery cluster;

[0018] multiplying the current value by a constant voltage value in a voltage reference source chip;

[0019] Integrating the multiplication result gives the charge and discharge capacity including the voltage coefficient.

[0020] The metering chip collects the current value of the total positive circuit of the entire battery cluster, including the following steps:

[0021] A shunt is connected in series in the total positive circuit of the entire battery cluster, and both ends of the shunt are connected to two current channel analog input ends of the metering chip;

[0022] Collect the voltage drop across the shunt;

[0023] Dividing the voltage drop by the resistance of the shunt yields the total positive circuit current of the entire battery cluster.

[0024] Generating a single battery power matrix according to the charge and discharge amount containing the voltage coefficient includes the following steps:

[0025] Eliminating the voltage coefficient in the charge-discharge capacity containing the voltage coefficient to obtain a heavy discharge capacity with the voltage coefficient eliminated;

[0026] A single cell power matrix is generated according to the arrangement of the batteries in the battery pack, wherein the rows and columns correspond to the batteries in the battery pack, and the values are the charge and discharge amounts for eliminating the voltage coefficient.

[0027] The step of eliminating the voltage coefficient in the charge and discharge quantity containing the voltage coefficient is specifically as follows:

[0028] The charge and discharge amount including the voltage coefficient is divided by the constant voltage value in the voltage reference source chip.

[0029] Generating the temperature sampling point distribution matrix includes the following steps:

[0030] Read the temperature value of each sampling point in the battery pack;

[0031] A temperature sampling point distribution matrix is generated according to the arrangement of the temperature sampling points in the battery pack, wherein the rows and columns correspond to the sampling points and the values thereof correspond to the temperature values.

[0032] The method of expanding the single cell power matrix and the temperature sampling point distribution matrix into matrices with the same rows and columns is specifically as follows:

[0033] Distributed calculation of the least common multiple of the rows and columns of the single cell power matrix and the temperature sampling point distribution matrix;

[0034] Amplifying the single battery power matrix using rows and columns of the least common multiple;

[0035] The temperature sampling point distribution matrix is amplified using the rows and columns of the least common multiple.

[0036] The amplification method is:

[0037] When performing row amplification, m rows are inserted between each row, and the value of each amplification multiple row is the value of the previous row, where m = row amplification multiple - 1;

[0038] When performing column amplification, n columns are inserted between each column, and the value of each amplification fold column is the value of the previous column, where n = column amplification fold - 1.

[0039] The present invention has the following advantages and beneficial effects:

[0040] 1. To address the instability caused by the MCU's use of a timer to perform ampere-hour integration, a separate energy metering chip is used to perform this integration task. Because metering chips are commonly used in household electricity meters, they have an extremely high sampling frequency. With a built-in 4MHz crystal oscillator, they can provide a 500kHz sampling frequency, significantly exceeding the sampling density of traditional BMSs that rely on MCU sampling. Furthermore, the metering chip's internal automatic integration significantly reduces MCU computing power. The MCU only needs to read the values in the metering chip's registers on a regular basis, significantly improving current integration accuracy.

[0041] 2. A dedicated chip is used for integration, unaffected by other tasks, interrupted by other tasks, or affected by MCU load. The metering chip has a built-in crystal oscillator that generates a very uniform clock signal and is immune to interference from other tasks. Even if the MCU reads data imprecisely for various reasons, the value read remains the exact value calculated by the metering chip, eliminating the influence of the MCU itself. Furthermore, through a rational layout, the metering chip is isolated from the main circuit, preventing interference from main circuit heat generation, and providing extremely accurate integration values.

[0042] 3. The metering chip itself has the function of counting the total charge and discharge amount of the system. It is equivalent to an electric meter and can provide quite accurate statistical data.

[0043] 4. Because the metering chip collects statistical energy data, it simultaneously collects real-time voltage and current data from the circuit, multiplies them, and then integrates them. The resulting real-time energy is different from the desired SOC data and cannot be directly used. Therefore, it is necessary to abandon the voltage collection function and design a dedicated circuit to provide a stable voltage source to the voltage collection terminal of the metering chip, so as to easily eliminate the voltage influence in the calculation.

[0044] 5. For the collected Anshi integral data and temperature data, a matrix is generated according to the battery arrangement. Through matrix amplification, one-to-one corresponding temperature values are obtained. Then, through the Kalman filter algorithm, the accurate SOC value is obtained. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is a circuit diagram of a metering chip according to an embodiment of the present invention;

[0046] Figure 2 Flowchart of a method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0047] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0048] Figure 1This is the schematic diagram of the metering chip circuit. U1 is the metering chip. Pin 1 is the positive power supply, connected to VCC. Pins 2 and 3 are the current channel analog inputs, connected in parallel across shunt R1. Shunt R1 is a low-value, high-precision resistor connected in series with the total positive circuit of the entire battery cluster. When current flows through the battery cluster, a certain voltage drop occurs across the shunt. The metering chip detects this voltage drop and converts it into the current value in the circuit. For example, R1 in the figure is a 0.006Ω shunt. When a current of 100A flows through the circuit, a 0.6V voltage drop occurs across R1. Pins 2 and 3 of U1 detect this 0.6V voltage drop and convert it into the current in the circuit. Pin 4 of U1 is the voltage input pin. Under normal conditions, the metering chip needs to monitor the voltage and current in the circuit in real time, multiplying and integrating them to obtain the energy in the circuit. However, in a lithium battery energy storage system, the total voltage of the battery cluster changes in real time during charging and discharging. Since calculating SOC only requires considering the relationship between current and time (Ah) in the circuit, real-time voltage sampling would hinder the calculation. Therefore, a voltage reference chip, U2, is used here to generate a constant voltage that is supplied to pin 4 of U1. Since the real-time current collected by the metering chip is multiplied by a constant voltage and then integrated, subsequent data processing only requires removing the constant voltage from the data to obtain the charge and discharge capacity. U2 is a constant voltage source. When pins 1 and 3 are connected, it generates a constant 2.5V reference voltage at pin 1. Pin 3 of U2 is positive and connected to current-limiting resistor R2, with a resistance of 1kΩ, which limits the current in the circuit to within U2's operating range. The other end of R2 is connected to VCC. Pin 2 of U2 is negative and connected to the circuit GND. Pin 1 of U2 is the reference pin, generating a 2.5V reference voltage and connecting to R3 and R4. R3 and R4 are voltage-divider resistors, with R3 being 24kΩ and R4 being 1kΩ. They divide the 2.5V generated by U2, generating a 0.1V voltage across R4, which serves as a constant input to U1's pin 4. U1's pin 5 is the chip ground and is connected to GND. U1's pin 6 is a function output pin, unused here and connected to GND. U1's pin 7 selects the communication mode: a high level indicates SPI communication, and a low level indicates UART communication. SPI communication is selected here, and it is connected to current-limiting resistor R5. R5 has a resistance of 1kΩ, and its other end is connected to VCC. U1's pins 8, 9, and 10 are communication pins, here for SPI communication. They are the clock, read, and transmit pins, respectively, and are connected to the MUC.

[0049] Figure 2The following is a system flow chart. When the energy storage system begins charging and discharging, the MCU sends a start signal to the metering chip. Since the metering chip collects energy data independently of the MCU, it is not affected by the MCU's operating state. When the metering chip begins operating, it simultaneously collects the current and voltage values on the shunt (the metering chip collects both voltage and current, but real-time voltage is not required here, so a reference voltage source is used instead). The chip multiplies the voltage and current values, integrates and accumulates them, and calculates the real-time electrical work. However, since only the charge and discharge capacity, not the electrical work, is of interest here, the influence of varying voltage should be excluded. This circuit uses a reference voltage source to generate a constant voltage signal, which can be removed during subsequent data processing. For example, if the metering chip collects a current of 100A and a constant voltage of 2.5V, it first multiplies the two to obtain an instantaneous power of 250W. It then integrates the instantaneous power, multiplies it by the inverse of the acquisition frequency, and adds it to the previous value to convert it into electrical work (Wh). For example, if the acquisition frequency is one hour, the result is 250Wh. At this time, the MCU reads the metering chip register and obtains the electric power data. However, since the voltage value is a constant 2.5V, the MCU only needs to divide it by the voltage value to obtain the power Ah required for SOC calculation, and the calculated value is 100Ah.

[0050] The MCU begins operation according to the acquisition timing and sampling frequency defined by the BMS system, for example, reading the metering chip data every 50ms. Because the MCU may be affected by various factors, such as the environment, task pressure, and internal interrupts, the resulting cycle is not strictly equal. For example, when the MCU is under low pressure, it can accurately measure time in 50ms, but under high pressure, it may only measure time once every 53ms. However, since the metering chip is independent of the MCU, the MCU's status has no effect on power statistics. The MCU reads the metering chip value at a predetermined interval and divides it by the voltage coefficient (the voltage is replaced by a reference voltage source, a fixed value of 2.5V, which is then divided down to 100mV and input to the metering chip) to obtain an accurate real-time power level. The MCU also reads the temperature value at each sampling point within the battery pack. Since the temperature sampling points do not correspond one-to-one with each battery cell, these temperature values cannot be directly used in SOC calculations. For example, if the batteries are arranged in 8 rows and 2 columns, the temperature sampling points should be arranged in 4 rows and 3 columns. Generate a single cell power matrix based on the arrangement of the batteries in the battery pack (defined according to the specific arrangement of the project batteries, such as 8 rows and 2 columns, 4 rows and 4 columns, etc.). Its rows and columns reflect the battery arrangement, and the values are the real-time power values read by the MCU. For example, an 8-row, 2-column matrix has values of 100Ah. Generate a temperature sampling point distribution matrix based on the arrangement of the temperature sampling points in the battery pack. Its rows and columns reflect the arrangement of the temperature sampling points, and the values are the real-time temperature values read. For example, a 4-row, 3-column matrix has values of actual temperature values. Amplify the two matrices according to the least common multiple of their rows and columns to generate two amplification matrices. At this time, the number of rows and columns of the two matrices is equal, and their values can be operated one-to-one (Kalman filter operation). For example, all are amplified to 8 rows and 6 columns. When performing row amplification, m rows are inserted between each row, and the value of each amplification multiple row is the value of the previous row, where m = row amplification multiple - 1; when performing column amplification, n columns are inserted between each column, and the value of each amplification multiple column is the value of the previous column, where n = column amplification multiple - 1. For example, the three rows (1, 2, 3) are expanded to six rows (1, 1, 2, 2, 3, 3). Using the Kalman filter algorithm for SOC calculation, the values of the two matrices are calculated one-to-one to obtain the accurate single-cell SOC value corrected by the Kalman filter.

[0051] An independent metering chip is used to calculate the charge and discharge values of the battery cluster in the lithium-ion battery energy storage system, avoiding the inaccurate SOC calculation caused by unstable MCU operating conditions and long sampling cycles. However, since the metering chip only calculates the real-time electrical work in the circuit, which is not what we need, a voltage reference chip is used to generate a constant voltage input to the metering chip. This ensures that the metering chip only calculates the changing current value in its internal calculations, and the constant voltage value can be easily removed in subsequent calculations.

[0052] Metering chips can provide extremely accurate power statistics, but they don't fully reflect the battery's SOC, which is often affected by temperature. However, the temperature sampling points in a battery pack often don't correspond one-to-one with the cells. Therefore, it's necessary to create a matrix based on their arrangement and then amplify the matrix to ensure a one-to-one correspondence between temperatures and cells. The Kalman algorithm is then used to calculate the precise SOC value.

Claims

1. A method for calculating SOC of a BMS system based on a metering chip, characterized in that: The metering chip is connected to the MCU and the voltage reference source chip, and the method includes the following steps: The MCU sends a start measurement command to the measurement chip, and the measurement chip calculates the charge and discharge amount including the voltage coefficient after receiving the start measurement command; The MCU reads the charge and discharge amount including the voltage coefficient calculated by the metering chip according to the acquisition timing defined by the BMS; Generate a single cell power matrix based on the charge and discharge quantities containing the voltage coefficient; and at the same time, generate a temperature sampling point distribution matrix; Expanding the single cell power matrix and the temperature sampling point distribution matrix into matrices with identical rows and columns; The two amplified single battery power matrices and temperature sampling point distribution matrices are subjected to Coleman filtering in a one-to-one correspondence to obtain the single battery SOC value; The calculation of the charge and discharge capacity including the voltage coefficient comprises the following steps: The metering chip collects the current value of the total positive circuit of the entire battery cluster; multiplying the current value by a constant voltage value in a voltage reference source chip; Integrate the multiplication result to obtain the charge and discharge capacity including the voltage coefficient; Generating a single battery power matrix according to the charge and discharge amount containing the voltage coefficient includes the following steps: Eliminating the voltage coefficient in the charge and discharge quantity containing the voltage coefficient to obtain the charge and discharge quantity with the voltage coefficient eliminated; A single cell power matrix is generated according to the arrangement of the batteries in the battery pack, wherein the rows and columns correspond to the batteries in the battery pack, and the values are the charge and discharge amounts for eliminating the voltage coefficient.

2. The method for calculating SOC of a BMS system based on a metering chip according to claim 1, characterized in that: The metering chip collects the current value of the total positive circuit of the entire battery cluster, including the following steps: A shunt is connected in series in the total positive circuit of the entire battery cluster, and both ends of the shunt are connected to two current channel analog input ends of the metering chip; Collect the voltage drop across the shunt; Dividing the voltage drop by the resistance of the shunt yields the total positive circuit current of the entire battery cluster.

3. The method for calculating SOC of a BMS system based on a metering chip according to claim 1, characterized in that: The step of eliminating the voltage coefficient in the charge and discharge quantity containing the voltage coefficient is specifically as follows: The charge and discharge amount including the voltage coefficient is divided by the constant voltage value in the voltage reference source chip.

4. The method for calculating SOC of a BMS system based on a metering chip according to claim 1, characterized in that: Generating the temperature sampling point distribution matrix includes the following steps: Read the temperature value of each sampling point in the battery pack; A temperature sampling point distribution matrix is generated according to the arrangement of the temperature sampling points in the battery pack, wherein the rows and columns correspond to the sampling points and the values thereof correspond to the temperature values.

5. The method for calculating SOC of a BMS system based on a metering chip according to claim 1, characterized in that: The method of expanding the single cell power matrix and the temperature sampling point distribution matrix into matrices with the same rows and columns is specifically as follows: Distributed calculation of the least common multiple of the rows and columns of the single cell power matrix and the temperature sampling point distribution matrix; Amplifying the single battery power matrix using rows and columns of the least common multiple; The temperature sampling point distribution matrix is amplified using the rows and columns of the least common multiple.

6. The method for calculating SOC of a BMS system based on a metering chip according to claim 5, characterized in that: The amplification method is: When performing row amplification, m rows are inserted between each row, and the value of each amplification multiple row is the value of the previous row, where m = row amplification multiple - 1; When performing column amplification, n columns are inserted between each column, and the value of each amplification fold column is the value of the previous column, where n = column amplification fold - 1.

Images