Abnormality detection device of secondary battery

Abstract

To provide a control system of secondary battery that performs abnormality detection while predicting other parameters, such as internal resistance and SOC with high accuracy.SOLUTION: A Kalman filter is used to detect a difference between an observed value (voltage) at a given time point and a voltage estimated using a prior state variable. A threshold voltage is set in advance and sudden abnormality, in particular micro short circuit, is detected by the detected difference voltage. In addition, neural network is used to learn data of time-series difference voltage. It is preferably determined and detected whether it is abnormal or normal.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] One aspect of the present invention relates to an article, a method, or a manufacturing method. Alternatively, the present invention relates to a process, Pertaining to a machine, manufacture, or composition of matter. One embodiment of the present invention is a semiconductor device, a display device, a light-emitting device, a power storage device, a lighting device, or an electronic Further, one embodiment of the present invention relates to a method for controlling charging of a power storage device, The present invention relates to a method for estimating a state of a power storage device and a method for detecting an abnormality in the power storage device, in particular, to a charging system for the power storage device, The present invention relates to a system for estimating a state of an electricity storage device and a system for detecting an abnormality in an electricity storage device.

[0002] In this specification, the term "electricity storage device" refers to elements and devices having an electricity storage function in general. For example, lithium ion secondary batteries and other storage batteries (also called secondary batteries) ion capacitors, nickel-metal hydride batteries, all-solid-state batteries, and electric double-layer capacitors. include.

[0003] Another embodiment of the present invention is a system for detecting an abnormality in a power storage device using a neural network. Another embodiment of the present invention relates to a neural network. Another aspect of the present invention relates to a vehicle using a neural network. In addition, one aspect of the present invention is not limited to a vehicle, but is also applicable to a device installed in a structure or the like. It can also be used in storage devices to store electricity generated by power generation facilities such as solar panels. This paper relates to a state estimation system and an equipment anomaly detection system. [Background technology]

[0004] In recent years, various types of storage devices such as lithium-ion secondary batteries, lithium-ion capacitors, and air batteries have become available. The development of lithium-ion batteries, which have high power output and high energy density, is particularly The secondary battery is used in mobile phones, smartphones, tablets, and notebook computers. Mobile information terminals, game devices, portable music players, digital cameras, medical equipment, hybrid Hybrid electric vehicle (HEV), electric vehicle (EV), or plug-in hybrid vehicle (PHEV) Next-generation clean energy vehicles such as electric motorcycles, etc., will be developed in line with the development of the semiconductor industry. Demand for rechargeable energy has expanded rapidly, and it is essential to today's information society as a source of rechargeable energy. It has become indispensable.

[0005] In addition, electric vehicles are vehicles that are driven only by electric motors, but they are also vehicles that use internal combustion engines such as engines. There are also hybrid vehicles that have both a battery and an electric motor. Several batteries are combined into one battery pack, and multiple sets of battery packs are installed under the car. is.

[0006] Secondary batteries used in electric vehicles, hybrid vehicles, and electric motorcycles have a high recharge rate and a low discharge depth. Deterioration occurs due to factors such as charging current and the charging environment (temperature changes). It depends on the temperature during charging, the frequency of quick charging, the amount of charge by regenerative braking, and The timing of charging the battery may also be a factor in deterioration. The secondary batteries used in hybrid vehicles are at risk of developing abnormalities such as short circuits due to deterioration over time. There is a problem.

[0007] In addition, secondary batteries used in electric vehicles, hybrid vehicles, and electric motorcycles are designed to be used for long periods of time. Since the premise is that the system will be used for a long time, it is desirable for the system to have sufficient reliability.

[0008] The remaining capacity (RC) of the lithium-ion secondary battery is the full charge capacity of the design capacity (DC). The percentage of the total capacity (FCC (Full Charge Capacity)), i.e. the charging rate The SOC is not set to use the entire capacity from 0% to 100%, and it is necessary to prevent over-discharge. To prevent this, a margin of 5% (or 10%) is provided instead of 0%. To prevent this, a margin of 5% (or 10%) is used instead of 100%, and as a result, Use the product within 5% to 95% (or 10% to 90%) of its design capacity. In reality, the BMS (Battery Management System) connected to the secondary battery is The upper limit voltage V max and the lower limit voltage V min Voltage range of By setting the value, the design capacity can be reduced by 5% to 95% (or 10% to 90%). ) for use.

[0009] Secondary batteries deteriorate with use, aging, and temperature changes. By accurately knowing the state of the battery, especially the SOC (state of charge), the secondary battery can be managed. This means that the upper limit voltage V max and the lower limit voltage V min It is also possible to widen the voltage range. The SOC was estimated by the coulomb counting method.

[0010] Patent Document 1 shows an example of using a neural network to calculate the remaining capacity of a secondary battery. It has been done. [Prior art documents] [Patent documents]

[0011] [Patent Document 1] U.S. Patent Publication No. 2006 / 0181245 Summary of the Invention [Problem to be solved by the invention]

[0012] Detects abnormalities in secondary batteries, for example, early detection of phenomena that reduce the safety of secondary batteries, and The company will warn users or change the operating conditions of the secondary battery to ensure safety. This is one of the challenges.

[0013] In addition, in conventional secondary battery abnormality detection, when the secondary battery deteriorates and an error occurs, correction is required. However, there is no feedback correction and it is insufficient, so the accuracy is low. One of the challenges is to increase

[0014] In addition, if a large amount of noise occurs in the secondary battery, the internal resistance and SOC of the secondary battery are monitored. When filtering is performed, the input noise data will cause errors in the SOC value that is estimated later. Ideally, while detecting abnormalities, other parameters (internal resistance, SOC, etc.) One of the challenges facing the company is developing a control system for secondary batteries that can predict the capacity of secondary batteries with high accuracy. [Means for solving the problem]

[0015] For lithium-ion batteries, only current, voltage and temperature parameters can be measured, and internal resistance and S OC (Charge Rate) is difficult to measure directly. Therefore, a regression model (regression equation) is used. For example, the internal resistance and SOC can be calculated by regression analysis, Kalman filter, or multiple regression analysis. Estimate.

[0016] The Kalman filter is a type of infinite impulse response filter. It is a type of variate analysis in which multiple independent variables are used in regression analysis. Regression analysis requires a large number of time series of observations, while the With the Mann filter, as long as a certain amount of data is accumulated, optimal correction coefficients can be obtained sequentially. In addition, the Kalman filter can be applied to non-stationary time series.

[0017] As a method for estimating the internal resistance and SOC of a secondary battery, a nonlinear Kalman filter (specifically For this, we can use the Unscented Kalman Filter (also called UKF). A Kalman filter (also called EKF) can also be used.

[0018] It is known that the internal resistance and SOC of a secondary battery can be estimated using a Kalman filter. However, it is difficult to detect sudden abnormalities, specifically micro-short circuits, using only this method. When estimating the internal resistance and SOC of a secondary battery, it is necessary to output a post-event state estimate. However, in the present invention, the state estimate is not used directly, but the difference between the observed value and the prior state estimate is used. By using this, it is possible to detect sudden abnormalities.

[0019] In order to solve the above problems, the present specification discloses an abnormality detection device and an abnormality detection system for a secondary battery. The system and anomaly detection method use the following means.

[0020] Using a Kalman filter, it is estimated using the observation value (voltage) at a certain point in time and the prior state variables. The difference between the detected voltage and the threshold voltage is detected. Detects sudden abnormalities, specifically micro-shorts, etc. By doing so, abnormalities in the secondary battery can be detected early.

[0021] A micro short circuit refers to a tiny short circuit inside a secondary battery. The positive and negative electrodes are not short-circuited to the point that charging and discharging becomes impossible, but rather, the short circuit is very small. This refers to the phenomenon in which a short circuit current flows suddenly. When the charge is repeated multiple times, the positive electrode active material is unevenly distributed, causing a part of the positive electrode and a part of the negative electrode Localized current concentration occurs in some parts of the separator, causing parts of the separator to stop functioning. It is said that micro-short circuits occur due to the generation of by-products from side reactions.

[0022] In an ideal secondary battery, the separator needs to be made thinner in order to reduce the size of the secondary battery. Furthermore, there is a demand for rapid charging at high voltages, and both of these have the potential to affect secondary batteries. This structure is prone to micro-short circuits. This can lead to serious accidents such as abnormal heat generation and fire in the secondary battery.

[0023] Therefore, in order to detect the occurrence of a micro-short circuit early and prevent a serious accident, Anomaly detection system, secondary battery control system, or secondary battery charging system Microshort circuits are an abnormality specific to secondary batteries, and in the past, attention was focused on microshort circuits. However, there has been no method or system for detecting micro-short circuits. A method for finding and calculating the value that changes significantly when a micro-short occurs In addition, we will develop AI (Artificial Intelligence) feedback correction using a neural network is inserted to detect any abnormalities in the secondary battery.

[0024] The measurement model for detecting abnormality in secondary batteries is shown below. This is a model of an anomaly detection system that performs a predetermined procedure on input to the system. The output from the system is determined by performing calculations or simulations using the By using techniques such as regression and learning, the optimal output can be determined for a given system input. Mechanisms for this purpose (e.g., neural networks, hidden Markov models, polynomial function approximations, etc.) ) are used as models. These models are merely examples and are not limiting.

[0025] The pre-estimation prediction step uses the model and input values, and the post-estimation step (filtering The step (also called the iteration step) uses observed values.

[0026]

number

[0027] The above equation is a state equation that describes the transition of the state of the system.

[0028] At a certain point in time (time k), the observed value y(k) is related to x(k) as follows:

[0029]

number

[0030] c T is the observation model that linearly maps the state space to the observation space. It is noise. The above equation is the observation equation.

[0031] The state equation and the observation equation are collectively called the state space model.

[0032] Moreover, the prior state estimate (left side) can be expressed by the following equation.

[0033]

number

[0034] Here, k is an integer such as 0, 1, 2, etc., and k is time. u(k) is the input signal and the secondary current In the case of a pond, it is the current value, and x(k) represents the state variable.

[0035] In addition, the prior error covariance (the left side of P - (k is the inverse of the covariance matrix) is given by the following formula: It can be expressed as:

[0036]

number

[0037] In the pre-estimation prediction step, the pre-state estimate and the pre-state covariance matrix are calculated based on the state equation. The posterior state estimate at time k and the posterior covariance matrix of the state and the state equation are calculated as Based on this, a prior state estimate and a prior covariance matrix at time k+1 are calculated.

[0038] The estimated value is compared with the actual voltage (observed value), and the error weighting coefficient is calculated using a Kalman filter. The Kalman gain is calculated to correct the estimate. The Kalman gain g(k) can be expressed as follows:

[0039]

number

[0040] The posterior state estimate (left side) used in the filtering step can be expressed as follows: do.

[0041]

number

[0042] In addition, the a posteriori error covariance matrix P(k) used in the filtering step can be expressed as follows: This can be done.

[0043]

number

[0044] According to the measurement model for detecting the occurrence of an abnormality in the secondary battery described above, the value of the following formula, i.e., Monitor the difference (voltage difference) between the observed value (voltage) at a point and the voltage estimated using the prior state variables If the behavior of the value changes significantly, it is assumed that an abnormality such as a micro short has occurred. This is how it is detected.

[0045]

number

[0046] When the differential voltage value of the above formula exceeds a certain threshold, a comparator or other device outputs a signal to detect an abnormality. When an abnormality is detected, a signal to display an abnormality to the outside or a switch The peaker outputs a signal to warn the user, such as a buzzer. The terms "detection" and "analysis" are used differently. "Detection" refers to detecting abnormal data and determining if the abnormal data is correct. The act of communicating with the outside, i.e. outputting a signal to another circuit, is called detection. "Detection" refers only to picking up abnormal data, not noise (incorrect abnormal data). Therefore, "detection" is a part of "detection", but is not equal to it. In addition, "detection" includes at least notification (signal output) to other circuits.

[0047] Also, when switching from a charging state to a discharging state, or from a discharging state to a charging state At times, the voltage difference fluctuates greatly, generating noise. This noise can lead to abnormalities in the secondary battery. Since this is not a phenomenon that occurs in a linear manner, multiple comparators may be provided to eliminate this noise.

[0048] The configuration of the invention disclosed in this specification includes a first detector for detecting a voltage value of a secondary battery, which is a first observation value. a detection means for detecting a current value of the secondary battery, which is a second observation value; A calculation unit that calculates a prior state estimate (estimated voltage value) using a Kalman filter based on the formula: The difference between the voltage value of the first observation and the estimated voltage value obtained at the previous time is calculated, and the difference is within a certain threshold range. and a determination unit that determines that the secondary battery is abnormal (such as a micro-short circuit) when the It is an anomaly detection device.

[0049] In the above configuration, the judgment unit has one or more comparators. This can eliminate noise and reduce errors in abnormality detection.

[0050] In addition, a neural network is used to learn the time series of differential voltage data and determine whether it is abnormal or normal. In the above configuration, it is preferable to determine and detect whether the voltage value of the first observation value is equal to or smaller than the voltage value of the previous observation value. and a neural network configuration unit to which the difference between the estimated voltage value obtained at each instant and the estimated voltage value obtained at each instant is input.

[0051] Another configuration disclosed in this specification is an anomaly detection method for determining whether a secondary battery is abnormal. A pre-estimation method for outputting an estimated voltage value using a Kalman filter based on a state equation A prediction step and a filtering step that calculates the posterior state estimates and the posterior error covariance matrix. The anomaly detection method has the following features:

[0052] Another configuration disclosed herein is a computer that performs a Kalman filter based on a state equation. A calculation unit that calculates a prior state estimate (estimated voltage value) using a filter, and a voltage value of the observed value and The difference between the estimated voltage value obtained at the previous time is calculated, and if it exceeds a certain threshold range, the secondary battery is considered abnormal. This is a program for causing the device to function as a determination unit that determines that the device is normal.

[0053] Using the above anomaly detection device, the above method, or a computer that executes the above program, The computer can also be used to configure a secondary battery abnormality detection system. The components are a control device, a smartphone, and a notebook personal computer. The control unit includes a control unit, a storage unit, and an input / output unit. The control unit includes a CPU (or MPU, MCU ( The control unit includes a GPU (Graphics Processing Unit) can also be used. A chip that integrates a CPU and a GPU is called an APU (Accelerated Processor). It is also called the APU (Asynchronous Serial Processing Unit), and this APU chip can also be used. Alternatively, an IC incorporating an AI system (also called an inference chip) may be used. An IC incorporating this system is called a circuit (microprocessor) that performs neural network calculations. There are cases like this.

[0054] The memory unit includes a RAM, a ROM, and a HDD. The input / output unit includes an operation unit, a display unit, and a communication unit. The program is stored in a memory unit of a computer. The present invention is not limited to the above, and includes a program stored in a computer-readable storage medium. The computer may read and execute the program. The body may be a disk such as a CD-ROM, a magnetic tape, a USB memory, or a flash memory. In addition, the Internet, LAN (Local Area Network), etc. ), the above program is stored in a device connected to a connection line such as a wireless LAN, and The computer may read and execute programs from these connections.

[0055] Another configuration disclosed in this specification is an anomaly detection system for detecting a micro-short circuit. a first detection means for detecting a voltage value of the secondary battery which is a first observation value; A second detection means for detecting a current value of the secondary battery; and a Kalman filter based on a state equation. a calculation unit that calculates a prior state estimation value (estimated voltage value) using the data; and a voltage value of the first observation value. The difference between the estimated voltage value obtained at the previous time is calculated, and if it exceeds a certain threshold range, the secondary battery is abnormal. and a determination unit for determining whether the defect is due to a micro-short circuit or the like. It is an anomaly detection system that detects abnormalities in data.

[0056] The secondary battery abnormality detection system disclosed in this specification constantly or periodically monitors the secondary battery. The sampling period (and calculation period) can be set appropriately. The secondary battery abnormality detection system disclosed in this document can also be called a secondary battery monitoring system. In addition, temperature sensors, cameras, gas sensors, etc. are used to detect abnormalities in the external surface temperature of the secondary battery, If the abnormality detection system for secondary batteries also includes detection of abnormalities such as deformation, the system will be more reliable. It may also be possible to detect anomalies.

[0057] The prediction error that is judged to be abnormal is not input directly to the Kalman filter, but is instead judged to be normal. The prediction error is input. The internal resistance and SOC of the secondary battery are calculated without using the abnormal values. Improve the accuracy of the estimate.

[0058] Another aspect of the invention disclosed in this specification is a secondary battery state estimation device that estimates the state of charge of a secondary battery. A method for determining a battery state by acquiring observed data from a secondary battery and estimating a prior state using a regression model. The predicted error voltage Vd, which is the difference between the observed value and the prior state estimate, is calculated. The data is judged as noise based on whether the voltage Vd data exceeds a preset threshold. The data judged as noise is replaced by the average value of k data before the anomaly is detected. The regression model is replaced with the input and correction is performed to detect abnormalities in secondary batteries that continue to be detected even after noise is detected. This is a state estimation method.

[0059] Micro-short circuit problems can occur during charging. For example, if the battery is configured with only one battery, In this case, the current is controlled by the charger, so the apparent current value does not change during a micro short. However, when batteries are connected in parallel, the voltage change becomes smaller and more difficult to detect. In addition, since this voltage change is within the upper and lower voltage limits of the battery, a special detection mechanism is required. In addition, when a micro-short circuit occurs in parallel batteries, the internal Because the resistance is low, the amount of current flowing through the healthy battery is relatively small, while the amount of current flowing through the abnormal battery is large. Current will flow, which is dangerous. However, since the current of the entire assembled battery is maintained at a controlled value, it is difficult to detect an abnormality. Also, in the case of a general assembled battery configuration, it is common to monitor the voltage of each series stage, but it is difficult to monitor the current of all batteries due to cost and wiring complexity.

[0060] As shown in the flowchart in FIG. 14, when compared with the signal REF in the comparison circuit and found to be smaller, that is, when the value of Equation 8 becomes <REF, it is regarded as the occurrence of an abnormality such as a micro short circuit, and after detecting this abnormality, data of the prediction error is created. For example, the average of the normal prediction errors from 1 step to 4 steps before is input into the Kalman filter. Even after detecting an abnormality, the SOC can be accurately obtained. The advantage of the Kalman filter is that it can predict the remaining capacity with high accuracy, and it can be said that the remaining capacity can be predicted even if the initial remaining capacity is unknown.

[0061] Conventionally, there has been a problem that an error occurs in the estimated value before and after the occurrence of a micro short circuit, and a deviation from the true capacity value occurs. By removing the data caused by the occurrence of a micro short circuit and inputting normal values, the accuracy of the estimation result can be improved.

[0062] Therefore, by preventing the data that was the basis for abnormality detection from being used in the estimation after abnormality detection, it is possible to use the secondary battery until a micro short circuit repeatedly occurs after abnormality detection.

[0063] An estimation method for estimating the state of charge of a secondary battery is shown below. After detecting the occurrence of an abnormality in the secondary battery, continue to repeat the procedure of performing the estimation. For the estimation, means such as regression and learning are used. ​​​​​A mechanism that can determine the optimal output for a system's input (e.g., neural (such as neural networks, hidden Markov models, and polynomial function approximations) are used to perform learning. Since it is preferable to use large amounts of data and analysis for learning, It may also be implemented within a site on a computer or server appliance, in which case it may be implemented on one or more servers. The data collection and analysis can be automated or semi-automated with the cooperation of an operator. In addition, if a large amount of data has already been stored and analyzed and the results obtained, Incorporating these results into the system, specifically into the memory of a program or IC chip It is also possible to detect anomalies and estimate the charging state without using a server.

[0064] In addition, the present invention also applies to the case where power is supplied wirelessly to charge a secondary battery. It is also possible to use a secondary battery abnormality detection system that detects abnormalities in the battery. The methods for wireless transmission over long distances include the electromagnetic induction method and the magnetic resonance method. The magnetic induction method is the Qi standard. The magnetic resonance method is WiPow. The receiving coil receives power from the power transmitting device, and the receiving coil and the secondary An abnormality detection device can be installed between the ponds. If the abnormality detection device detects an abnormality, the power transmission device The instruction to stop power supply from the This is done using a Bluetooth (registered trademark).

[0065] The embodiments described herein below may be implemented using various computer hardware or software. This includes the use of special-purpose or general-purpose computers, including software. The embodiments described below in the specification are implemented using a computer-readable recording medium. The recording medium can be a RAM, a ROM, an optical disk, a magnetic disk, etc. or any other storage medium that can be accessed by a computer. Also, the algorithms shown as examples in the embodiments described below in this specification , components, flows, programs, etc. may be implemented in software or in hardware. The implementation may be in a combination of hardware and software. Effect of the Invention

[0066] By monitoring the value of the above formula 8 (differential voltage), it is possible to easily and accurately measure the difference between the secondary batteries. Furthermore, feedback correction is possible using a neural network. By using this function to detect abnormalities in the secondary battery, it is possible to detect abnormalities in the secondary battery with higher accuracy. become.

[0067] In addition, the system is not limited to detecting abnormalities in a single secondary battery, but can also detect abnormalities in multiple secondary batteries connected in series. It is also possible to perform anomaly detection on the data.

[0068] In addition, the secondary battery is not limited to a lithium-ion secondary battery using an electrolyte, but may be a battery using a solid electrolyte. All-solid-state batteries, sodium ion secondary batteries, potassium ion secondary batteries, etc. Potassium ion secondary batteries attract solvents more easily than lithium or sodium batteries. The force is weak, allowing ions to move freely through the electrolyte. The types and sizes of secondary batteries are changing. In this case, the threshold value is set appropriately according to the secondary battery. Since cross-shorts can occur, the anomaly detection system disclosed herein is useful.

[0069] The abnormality detection system disclosed in this specification is mounted on an IC chip or the like and is one of the systems of a vehicle. It can also be incorporated in other functional circuits (random access memory ( Random Access Memory (RAM), GPU (Graphics P Processing Unit), PMU(Power Management Unit) t) etc.) may be integrated into a single IC chip.

[0070] The anomaly detection system disclosed in this specification can shorten the detection timing and is real In addition, it is possible to detect abnormalities in a short time regardless of the state of the secondary battery, such as when charging or discharging. Therefore, abnormality detection can be achieved regardless of the

[0071] In addition, it detects abnormalities in secondary batteries in real time, removes noise used to detect abnormalities, and A secondary battery control system that predicts the parameters (internal resistance, SOC, etc.) with high accuracy It can be achieved. [Brief description of the drawings]

[0072] [Figure 1] 1 is an equivalent circuit model illustrating one embodiment of the present invention. [Diagram 2] FIG. 1A is a functional block diagram showing one embodiment of the present invention, and FIG. 1B is a graph showing the relationship between differential voltage and time. [Diagram 3] FIG. 1 is a functional block diagram illustrating an embodiment of the present invention. [Figure 4] FIG. 1 is a functional block diagram illustrating an embodiment of the present invention. [Diagram 5] FIG. 1 is a functional block diagram illustrating an embodiment of the present invention. [Figure 6] FIG. 1 is a functional block diagram illustrating an embodiment of the present invention. [Figure 7] FIG. 1 is a functional block diagram illustrating an embodiment of the present invention. [Figure 8] FIG. 13 is a graph showing a simulation result using a measurement model showing one embodiment of the present invention. [Figure 9] FIG. 13 is a graph showing a simulation result using a measurement model showing one embodiment of the present invention. [Figure 10] 1A and 1B are a block diagram and a perspective view of an electric vehicle and a secondary battery, illustrating one embodiment of the present invention. [Figure 11] FIG. [Figure 12] FIG. 1 is a perspective view showing an example of a secondary battery. [Figure 13] 1A and 1B are a cross-sectional view and a perspective view showing an example of a secondary battery. [Figure 14] 1 is a flow diagram showing one embodiment of the present invention. [Figure 15] FIG. 1 is a diagram showing an equivalent circuit model illustrating one embodiment of the present invention. [Figure 16] FIG. 1 is a system diagram illustrating one embodiment of the present invention. [Figure 17] FIG. 1 is a system diagram illustrating one embodiment of the present invention. [Figure 18] 1 is a flow diagram showing one embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0073] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. The present invention is not limited to the following description, and those skilled in the art will recognize that the present invention may be modified in various ways in form and detail. The present invention is not limited to the description of the following embodiments. It is not something that can be done.

[0074] (Embodiment 1) FIG. 1 is an example of an equivalent circuit model of a battery (also called a battery model) used for abnormality detection. In FIG. 1, R0 is a series resistance component, and R1, R2, R3, and R4 are resistors. C3 and C4 are capacitances.

[0075] A pulse signal is applied to the micro short model 400 in FIG. A cross-short circuit is artificially generated to perform a numerical simulation of the battery model. In this paper, we will explain the method using numerical simulations, but in reality, we will monitor the voltage of the battery with an abnormality. Observe.

[0076] The OCV shown in Figure 1 is an abbreviation for Open Circuit Voltage. This is the nominal value when the battery is disconnected from the external circuit for a sufficient period of time and the electrochemical reaction inside the battery This is the potential difference between the positive and negative electrodes when an equilibrium state is reached.

[0077] FIG. 2(A) is a diagram showing an example of a functional block, which can also be called estimation logic.

[0078] In FIG. 2A, a delay circuit 402 delays a state estimate at time k to an estimate at time (k+1). In Fig. 2(A), A and b are matrices obtained from the model. C T is the observation coefficient vector. Z -1 is a delay circuit.

[0079] In FIG. 2(A), the part indicated by reference numeral 401 is estimated using the observed value (voltage) and the prior state variable. If the differential voltage changes suddenly, a micro short circuit will occur. It can be considered that the voltage difference is generated, so the value of this voltage difference is input to the comparator 403. In 403, the reference signal (REF) is compared with the threshold value to determine whether it is abnormal or not. The comparator 403 compares the two input values ​​and outputs the lower one as LO. REF is changed several times within one step period and a multi-level comparison is performed. Good too.

[0080] Figure 2(B) shows the relationship between the terminal voltage of the secondary battery, the differential voltage, and time. The horizontal axis is time, and the vertical axis is time. The axis is voltage. The fluctuation in the terminal voltage is not so large, so the micro short circuit It is difficult to identify the occurrence timing. On the other hand, as shown in Figure 2(B), The fluctuations are reflected in the differential voltage (voltage difference between the observed value and the prior state estimate), and the micro-short circuit is Therefore, by monitoring the differential voltage, it is possible to detect the occurrence of a micro-short circuit. You can judge life.

[0081] In addition, the estimation logic shown in Figure 2(A) is partially shared using the same input data. By using this function, it is possible to output the SOC estimate. The estimate can be output.

[0082] (Embodiment 2) In this embodiment, a neural network (NN) is further used to improve accuracy. This is explained below.

[0083] FIG. 3 is a diagram showing an example of a functional block. NN in FIG. 3(A) is a neural network part. and the output is LO(k).

[0084] In FIG. 3(A), the part indicated by reference numeral 401 is estimated using the observed value (voltage) and the prior state variable. This is the difference between the measured voltage and the actual voltage (difference voltage), and the value of this difference voltage is input to the neural network. The data is input to 404 and compared with the learning data to determine whether it is abnormal or not. The data input to the neural network unit 404 is stored and used as part of the learning data. The accuracy can be further improved.

[0085] FIG. 3(B) shows another variation. The functional block shown in FIG. 3(B) is The difference voltage value is input to both the comparator and the neural network (NN), and the OR is used. Also, AND can be used instead of OR.

[0086] FIG. 4(A) shows another variation. The functional block shown in FIG. 4(A) is The output of the comparator is input to a neural network (NN) to determine whether the output value is abnormal. A determination is made as to whether or not

[0087] FIG. 4(B) shows another variation. The functional block shown in FIG. 4(B) is The output of the comparator is input to a neural network (NN), and the comparator and the neural network The value obtained by ANDing the work (NN) is used to determine whether or not there is an abnormality.

[0088] FIG. 5(A) shows another variation. The functional blocks shown in FIG. is configured to detect anomalies using data that has passed through the Kalman gain g(k).

[0089] (Embodiment 3) In this embodiment, a configuration for improving accuracy by using another comparator will be described below. Reveal.

[0090] Examine the relationship between the passage of time and the value of the differential voltage, and find that there is an error different from that when a microshort occurs. The present inventors discovered that an error occurs when switching between charging and discharging in a secondary battery. As shown in Figure 6(A), it is clear that the terminal current changes with the timing of discharge. do.

[0091] When switching between charging and discharging, no overshoot on the positive side is observed. As shown in 6(B), when a micro short occurs, the on-resistance is applied to both the positive and negative sides. A positive overshoot occurs. The correction of the forecast error is This is because the negative overshoot occurred first, and then the positive overshoot occurred. Due to this difference, overshoot occurs only on the positive side. In this case, it is determined as an error and the occurrence of a micro short circuit is not determined.

[0092] As shown in Figure 5(B), two comparators are provided, and comparator 1 is set to detect the negative overshoot. Input the value (REF1) of the positive overshoot to comparator 2 (REF2). The output of the comparator one step before and the current output of the comparator are input using a delay circuit, Perform D calculation.

[0093] The above configuration eliminates noise when switching between charging and discharging, and prevents micro-short circuit abnormalities. The detection accuracy can be improved.

[0094] This embodiment mode can be freely combined with other embodiment modes.

[0095] (Embodiment 4) In this embodiment, an example of application to an electric vehicle (EV) will be described with reference to FIG.

[0096] FIG. 10A shows an example of a block diagram of an electric vehicle.

[0097] The electric vehicle includes a first battery 301 as a secondary battery for main driving, and a motor 30 A second battery 311 is provided to supply power to an inverter 312 which starts the engine 4. In this embodiment, the abnormality monitoring unit 300 is driven by the power source of the second battery 311. The multiple secondary batteries constituting the first battery 301 are collectively monitored.

[0098] The first battery 301 mainly supplies power to the 42V (high voltage) in-vehicle devices, and the second The battery 311 supplies power to the 14V system (low voltage system) in-vehicle devices. 11 Lead-acid batteries are often used because of their cost advantage. Lead-acid batteries are not compatible with lithium-ion batteries. Compared to secondary batteries, they have a higher self-discharge rate and are more susceptible to deterioration due to a phenomenon called sulfation. By using a lithium ion secondary battery as the second battery 311, maintenance is simplified. However, when used for a long period of time, for example, for more than three years, There is a risk of an abnormality occurring that cannot be detected. When the first battery 301 becomes inoperable, the motor is not started even if the first battery 301 has remaining capacity. In order to prevent this, if the second battery 311 is a lead-acid battery, The battery supplies power to a second battery, which is then charged to keep it fully charged at all times. is.

[0099] In this embodiment, both the first battery 301 and the second battery 311 contain lithium ions. The second battery 311 is a lead-acid battery or a solid-state battery. Good too.

[0100] An example of a cylindrical secondary battery will be described with reference to Figs. 12(A) and 12(B). As shown in FIG. 12(A), the secondary battery 600 has a positive electrode cap (battery lid) 60 on the top surface. 1, and a battery can (outer can) 602 on the side and bottom. and the battery can (external can) 602 are insulated by a gasket (insulating packing) 610. There are.

[0101] Fig. 12(B) is a schematic diagram showing a cross section of a cylindrical secondary battery. Inside the can 602, a strip-shaped positive electrode 604 and a negative electrode 606 are sandwiched between a separator 605. A wound battery element is provided. Although not shown, the battery element is wound around a center pin. The battery can 602 is closed at one end and open at the other end. The material is nickel, aluminum, titanium, or other metals that are resistant to corrosion by the electrolyte. These and their alloys with other metals (e.g., stainless steel, etc.) can be used. In addition, in order to prevent corrosion by the electrolyte, it is preferable to coat the electrode with nickel, aluminum, or the like. A battery element in which a positive electrode, a negative electrode, and a separator are wound inside a battery can 602. The battery element is sandwiched between a pair of opposing insulating plates 608 and 609. A non-aqueous electrolyte (not shown) is injected into the battery can 602. Lithium baltic oxide (LiCoO 2 ) and lithium iron phosphate (LiFePO 4 ) and other active materials a positive electrode including a positive electrode and a negative electrode made of a carbon material such as graphite capable of absorbing and releasing lithium ions; In organic solvents such as ethylene carbonate and diethyl carbonate, LiBF 4 or LiPF 6 The lithium salt may be a non-aqueous electrolyte solution.

[0102] The positive and negative electrodes used in cylindrical secondary batteries are wound, so active material is formed on both sides of the current collector. A positive electrode terminal (positive electrode current collecting lead) 603 is connected to the positive electrode 604, and a negative A negative electrode terminal (negative electrode current collecting lead) 607 is connected to the positive electrode 606. The positive terminal 607 may be made of a metal material such as aluminum. 03 is resistance welded to the safety valve mechanism 612, and the negative terminal 607 is resistance welded to the bottom of the battery can 602. The safety valve mechanism 612 is a PTC (Positive Temperature Coefficient) It is electrically connected to the positive electrode cap 601 via a passive element 611. The safety valve mechanism 612 is a mechanism for releasing the positive electrode cap 601 when the internal pressure of the battery increases beyond a predetermined threshold. The PTC element 611 cuts off the electrical connection between the positive electrode 604 and the positive electrode 604. It is a thermal resistor element whose resistance increases when the temperature rises, and the amount of current is limited by the increase in resistance. It prevents abnormal heat generation. The PTC element is made of barium titanate (BaTiO 3 )system Semiconductor ceramics and the like can be used.

[0103] A lithium ion secondary battery using an electrolyte includes a positive electrode, a negative electrode, a separator, an electrolyte, and In lithium-ion secondary batteries, the anode (positive electrode) is charged and discharged. The cathode (negative electrode) is switched, and the oxidation reaction and reduction reaction are switched, An electrode with a high reaction potential is called a positive electrode, and an electrode with a low reaction potential is called a negative electrode. In the detailed description, even if a reverse pulse current is applied during charging or discharging, Whether it is a battery that carries a current or a battery that carries a charging current, the positive electrode is called the "positive electrode" or the "+ electrode (plus electrode)". The negative electrode is called the "negative electrode" or "-electrode (minus electrode)". The terms anode and cathode, which are related to the reaction, are used during charging and discharging. At times, the anode is the opposite, which can be confusing. The terms cathode and cathode are not used in this specification. When using the terms anode and cathode, specify whether they are charging or discharging. Also indicate whether it corresponds to a positive electrode (plus electrode) or a negative electrode (minus electrode). do.

[0104] A charger is connected to the two terminals shown in FIG. 12(C) to charge the storage battery 1400. In FIG. 12(C), 1406 is an electrolyte and 1408 is a separator. As the charging of the battery 1400 progresses, the potential difference between the electrodes increases. From the external terminal, current flows toward the positive electrode 1402, and inside the storage battery 1400, The current flows from the negative electrode 1404 to the external terminal of the storage battery 1400. The direction of the current is defined as the positive direction. In other words, the direction of the current is defined as the positive direction.

[0105] In this embodiment, an example of a lithium ion secondary battery is shown, but the present invention is not limited to the lithium ion secondary battery. For example, a material having element A, element X, and oxygen may be used as a positive electrode material for a secondary battery. The element A is one or more selected from the group 1 elements and the group 2 elements. As the element of Group 1, for example, lithium, sodium, potassium, etc. The elements of Group 2 include, for example, calcium, beryllium, and potassium. The element X may be, for example, a metal element, silicon, magnesium, etc. and phosphorus. The element X can be cobalt, nickel, or the like. It is preferable that the metal is at least one selected from the group consisting of manganese, iron, and vanadium. , lithium cobalt oxide (LiCoO 2 ) and lithium iron phosphate (LiFePO 4 ) are mentioned.

[0106] The negative electrode has a negative electrode active material layer and a negative electrode current collector. The negative electrode active material layer contains a conductive assistant and and a binder.

[0107] As a negative electrode active material, it is possible to carry out charge / discharge reactions by alloying / de-alloying reactions with lithium. For example, silicon, tin, gallium, aluminum, and Rumanium, lead, antimony, bismuth, silver, zinc, cadmium, indium, etc. Materials containing at least one of these elements can be used. These elements have a large capacity compared to carbon. Silicon in particular has a high theoretical capacity of 4200mAh / g.

[0108] In addition, the secondary battery preferably has a separator. Examples of the separator include Cellulose-containing fibers, including paper, nonwoven fabrics, glass fibers, ceramics, or Nylon (polyamide), Vinylon (polyvinyl alcohol fiber), polyester, a Use synthetic fibers such as acrylic, polyolefin, and polyurethane. This can be done.

[0109] In addition, the regenerative energy generated by the rotation of the tire 316 is transmitted to the motor 304 via the gear 305. The motor controller 303 and the battery controller 302 control the second battery. 311 or the first battery 301 is charged.

[0110] The first battery 301 is mainly used to rotate the motor 304, but the DC Through the DC circuit 306, 42V system vehicle components (electric power steering 307, heater 308, In the case where the rear wheels have a rear motor, the first The battery 301 is used to rotate the rear motor.

[0111] The second battery 311 also supplies 14V-based in-vehicle components (such as The power supply 314 supplies power to various devices (such as a stereo 313, power windows 314, and lamps 315).

[0112] Moreover, the first battery 301 is composed of a collection of modules including a plurality of secondary batteries. For example, a cylindrical secondary battery 600 shown in FIG. 12(A) is used. As shown in FIG. 1, a cylindrical secondary battery 600 is sandwiched between conductive plates 613 and 614 to form a module. In FIG. 10(B), no switch is shown between the secondary batteries. The secondary batteries 600 may be connected in parallel, in series, or in parallel. After being connected in series, the secondary batteries 600 may be further connected in series. By constructing a joule, it is possible to extract a large amount of power.

[0113] In order to cut off the power from multiple secondary batteries in a vehicle, a high-speed It has a service plug or circuit breaker that can cut off the voltage and the first battery For example, 48 battery modules each having 2 to 10 cells are provided in the battery 301. When connecting directly, a service plug or circuit breaker must be inserted between the 24th and 25th terminals. It has a laker.

[0114] FIG. 11 illustrates a vehicle using a secondary battery abnormality detection system according to an embodiment of the present invention. The secondary battery 8024 of the automobile 8400 shown in FIG. It not only drives the headlights 8401 and room lights (not shown), but also drives other light-emitting devices. The secondary battery 8024 of the automobile 8400 can supply power to the The cylindrical secondary battery 600 shown in FIG. 1 is sandwiched between a conductive plate 613 and a conductive plate 614 to form a module. A module may also be used.

[0115] The automobile 8500 shown in FIG. 11(B) is a secondary battery that is plugged in. The device can be charged by receiving power from an external charging facility using a contactless charging method or other methods. FIG. 11(B) shows a diagram of a charging device 8021 mounted on a ground and a charging station 8022 mounted on a vehicle 8500. The secondary battery 8024 is shown being charged via a cable 8022. For charging methods and connector specifications, please refer to the designated specifications of CHAdeMO (registered trademark) or Combo. The charging device 8021 is a charging station installed in a commercial facility. For example, plug-in technology can be used to The secondary battery 8024 installed in the automobile 8500 can be charged by the power supply. Charging is performed by converting AC power to DC power via a conversion device such as an ACDC converter. It is possible.

[0116] Although not shown, a power receiving device is mounted on the vehicle and receives power from a ground power transmitting device in a non-contact manner. In this non-contact power supply method, a power transmission device is installed on the road or exterior wall. By incorporating this technology, charging can be done not only when the vehicle is stopped but also while it is moving. Using this method, electric power may be transmitted between vehicles. A solar battery may be installed to charge the secondary battery when the vehicle is stopped or running. The power can be supplied using an electromagnetic induction method or a magnetic resonance method.

[0117] FIG 11C shows an example of a two-wheeled vehicle using the secondary battery of one embodiment of the present invention. The scooter 8600 shown in (C) has a secondary battery 8602, side mirrors 8601, and a turn signal. The secondary battery 8602 can supply electricity to the direction indicator light 8603. can.

[0118] In addition, the scooter 8600 shown in FIG. 11C has a secondary battery 860 in the storage space under the seat 8604. 2 can be stored. The secondary battery 8602 can be stored in the under-seat storage 8604 even if it is small. , and can be stored in the under-seat storage space 8604.

[0119] The secondary battery 8602 can be an all-solid-state battery. The battery is constructed of a laminated type secondary battery. Here, a laminated type secondary battery using an all-solid-state battery is used. An example is shown in FIG.

[0120] The laminated secondary battery 500 shown in FIG. 13(D) has a positive electrode lead electrode 510 and a negative electrode lead electrode. The semiconductor device has a gate electrode 511.

[0121] The procedure for making a laminated secondary battery will be briefly described below. First, a positive electrode and a negative electrode are prepared. The electrode has a positive electrode current collector, and the positive electrode active material layer is formed on the surface of the positive electrode current collector. The positive electrode current collector has a region where the positive electrode current collector is partially exposed (hereinafter referred to as a tab region). The negative electrode active material layer is formed on the surface of the negative electrode current collector. It has a partially exposed area, i.e., a tab area.

[0122] Then, the negative electrode, the solid electrolyte layer, and the positive electrode are laminated. Here, five sets of negative electrodes and four sets of positive electrodes are used. Next, we will show an example of bonding the tab regions of the positive electrodes together and attaching a positive electrode lead to the tab region of the topmost positive electrode. The electrode 510 is then bonded. For example, ultrasonic welding or the like may be used for the bonding. Bonding of the negative electrode tab regions to each other and bonding of the negative electrode lead electrode 511 to the negative electrode tab region on the outermost surface Do the following.

[0123] Next, a negative electrode, a solid electrolyte layer, and a positive electrode are placed on the exterior body. Any material layer (such as ceramic) containing solid components capable of conducting lithium ions can be used. For example, The solid electrolyte layer is formed by forming a sheet from a slurry of ceramic powder or glass powder. The definition of ceramics is a material that is free of metals, nonmetals, oxides, carbides, nitrides, borides, etc. Glass is an organic compound material. Glass is defined as an amorphous material that exhibits the glass transition phenomenon. However, when made into a microcrystalline form, it is sometimes called ceramic glass. Because it has crystallinity, it can be confirmed by X-ray diffraction. A solid electrolyte such as a sulphide solid electrolyte or a sulfide solid electrolyte can be used. The active material layer also contains a solid electrolyte and may contain a conductive additive. Any material having conductivity may be used, for example, a carbon material, a metal material, etc. can.

[0124] In addition, the oxide solid electrolyte used as the positive electrode active material is Li 3 PO 4 , Li 3 B O 3 , Li 4 SiO 4 , Li 4 GeO 4 , LiNbO 3 , LiVO 2 , LiTiO 3 , L iZrO 3 In addition, a composite compound of these may be used, for example BaLi 3 BO 3 -Li 4 SiO 4 The surface of the solid electrolyte is At least a part of the substrate may be covered with a coating layer having a thickness of 1 nm or more and 20 nm or less. The material used is a Li-ion conductive oxide.

[0125] The oxide solid electrolyte used as the negative electrode active material is Nb 2 O 5 , Li 4 Ti 5 O 1 2 In this specification, SiO is, for example, silicon monoxide. Or SiO refers to SiO 2 This refers to materials that have a higher silicon content than SiO x Here, it is preferable that x has a value close to 1. For example, x is preferably 0.2 or more and 1.5 or less, and more preferably 0.3 or more and 1.2 or less.

[0126] In addition, the sulfide solid electrolyte used as the positive electrode active material includes a material containing Li and S, Specifically, Li 7 P 3 S 11 , Li 2 S-SiS 2 , Li 2 SP 2 S 5 Examples include: This can be done.

[0127] Next, the exterior body is folded. After that, the outer periphery of the exterior body is joined. The exterior body is made of metal foil and organic Laminate film made by laminating a resin film, such as aluminum foil or stainless steel foil In this way, the bonding is performed by, for example, thermocompression bonding. In this embodiment, a laminate type secondary battery 500 can be fabricated by using a single laminate. In the example shown here, two laminate films are used to bond the edges of the Alternatively, the sealing may be achieved by bonding the same to the substrate.

[0128] FIG. 13(A) is a conceptual diagram of a solid-state battery, in which a solid is disposed between a positive electrode 81 and a negative electrode 82. The solid-state battery has an electrolyte layer 83. In addition, there are thin-film type all-solid-state batteries and bulk type all-solid-state batteries. A thin-film all-solid-state battery is an all-solid-state battery obtained by stacking thin films. A silicon-type all-solid-state battery is an all-solid-state battery obtained by stacking fine particles.

[0129] FIG. 13B shows an example of a bulk-type all-solid-state battery, in which particulate positive electrode active material is disposed near the positive electrode 81. The negative electrode 82 has a particle-shaped negative electrode active material 88 in the vicinity of the negative electrode 82, and the particle-shaped negative electrode active material 88 is disposed so as to fill the gap between them. A solid electrolyte layer 83 is disposed between the positive electrode 81 and the negative electrode 82. A gap is formed between the positive electrode 81 and the negative electrode 82 by pressing. Multiple types of particles are packed to eliminate this.

[0130] FIG. 13(C) shows an example of a thin-film all-solid-state battery. Thin-film all-solid-state batteries are manufactured by a gas phase method ( Air deposition method, thermal spray method, pulsed laser deposition method, ion plating method, cold spray The film is formed by using a method such as deposition, aerosol deposition, or sputtering. After forming wiring electrodes 85 and 86 on a substrate 84, a positive electrode 81 is formed on the wiring electrode 85. A solid electrolyte layer 83 is formed on the positive electrode 81, and a negative electrode 8 is formed on the solid electrolyte layer 83 and the wiring electrode 86. This is an example in which a lithium ion battery is manufactured by forming a substrate 84 using a ceramic Examples of the substrate include a metal substrate, a glass substrate, a plastic substrate, and a metal substrate.

[0131] This embodiment mode can be appropriately combined with the descriptions of other embodiment modes.

[0132] (Embodiment 5) An example of a method for estimating the SOC of a secondary battery is shown in FIG. 14. FIG. 14 is a flow chart. After detecting an abnormality such as a micro-short, the prediction error data is created, for example, from one step The average of the normal prediction errors from the last four steps is input to the Kalman filter. Even after the procedure, the SOC can be accurately calculated.

[0133] When the differential voltage value of the above formula exceeds a certain threshold, a comparator or other device outputs a signal to detect an abnormality. The comparator detects abnormalities by comparing the threshold voltage signal REF with the comparator. The data at the time when the abnormality was detected is not used for the subsequent estimation, but instead, the data from several steps before is used. The average value of the difference voltage in the above formula 8 is input to the estimation algorithm. When the signal falls below REF, it is replaced with the average value of the previous several steps. Therefore, the above formula When the difference voltage of 8 falls below the voltage signal REF input to the comparator, the difference voltage becomes Karma The average is not fed into the filter loop. Instead, the average is fed into the estimation algorithm. This allows for highly accurate estimation of SOC even when an abnormality occurs. Instead of using the data at the time when the abnormality was detected, the average value for the previous few steps is used. If the above equation (8) is input into the estimation algorithm, the value of the differential voltage is The data is close to that when it is not used.

[0134] FIG. 16 shows a specific system diagram for executing the flow of FIG. 14. In FIG. 16, The secondary battery charge state estimating device includes at least a comparator 403, a delay circuit, and an AND circuit 404. 05 and a multiplexer 407. The clock signal CLK is input to the AND circuit. Also, a reference signal REF is input to the comparator 403. FIG. 16 is an example, and there is no particular limitation. The device for estimating a state of charge of a secondary battery is configured to estimate a state of charge of a secondary battery without determining a first observation value. A detection means for detecting a voltage value of the secondary battery and a regression model for calculating an estimated voltage value. A calculation unit calculates the difference between the voltage value of the first observation value and the estimated voltage value obtained at the previous time, and calculates a threshold. and a determination unit that determines that the secondary battery is abnormal when the detected value exceeds a certain range. Or, it has a plurality of comparators, a multiplexer, and a delay circuit. The secondary battery state of charge estimation device further obtains the second observation value of the secondary battery. A second detection means for detecting the current value may be provided. The system diagram is shown in Figure 17. In FIG. 17, an example of a different IIR (Infinite Impulse Response) filter is shown. In Fig. 17, N is a sufficiently large value of the time k. This refers to the case where a time limit is set.

[0135] In addition, even if the data that detects an anomaly is not input to the Kalman filter loop, the anomaly detection The SOC is calculated by accurately calculating and reflecting the current lost due to a micro-short circuit at the time of power output. This allows for more accurate values. Also, a flow chart is shown in Figure 18. In other words, the value of Equation 8 is expressed as LO. <R When it becomes EF, it is assumed that an abnormality such as a micro short has occurred, and this abnormality is detected.

[0136] The data that detects the abnormality is the predicted error voltage, and the state equation is used to calculate the voltage at the time of the micro short circuit. current I micro Using the equivalent circuit model shown in Fig. 15(A) and Fig. 15(B), In FIG. 15(A) and FIG. 15(B), OCV is the potential difference during discharge. Yes, V 0 , V 1 , V 2 , V 3 is the voltage at each point.

[0137]

number

[0138] The above equation is the state variable x(k) of the circuit in FIG. 15(A). is an equivalent circuit model corresponding to the state before the occurrence of a micro-short circuit.

[0139] Also, u(k) is the current I BAT (k). u(k) is the input signal and in the case of a secondary battery , the current value.

[0140]

number

[0141] b, which constitutes the state equation, is a constant, and T S is the sampling period.

[0142]

number

[0143] The above equation is the state equation of the Kalman filter. 1 , R 2 , R 3 , capacity C 1 , C 2 , C 3 , the full charge capacity FCC can also be expanded to include the state variable x(k) good.

[0144] Next, the state when a microshort occurs is considered as the equivalent circuit model shown in Figure 15(B). The calculation procedure is shown below.

[0145]

number

[0146] The above formula is a relational expression when the micro-short circuit occurs at time k+1. The current at can be shown as follows:

[0147]

number

[0148] In addition, the resistor R 1 and capacity C 1 The voltage V applied to 1 is expressed as follows:

number

[0149] In addition, the resistor R 2 and capacity C 2 The voltage V applied to 2 is expressed as follows:

[0150]

number

[0151] In addition, the resistor R 3 and capacity C 3 The voltage V applied to 3 is expressed as follows:

[0152]

number

[0153] If the values ​​shown in the above formula are much smaller than 1 or if you do not require high accuracy, , the value shown below may be set to 1.

[0154]

number

[0155] If the above formula is set to 1, the amount of calculation can be reduced.

[0156]

number

[0157] Calculate the above formula and calculate the current (I micro ) can be obtained. As shown in the formula, the R estimated one step before 0 , OCV, voltage including predicted error voltage V IN and current I BATThe current at the time of the micro-short circuit is calculated using the observed value of R 0 (k) is the covariance of the observation errors.

[0158]

number

[0159] In the above formula, SOC(k) is the SOC data inside the Kalman filter at the pre-estimation prediction step. The value on the left side of the above formula corresponds to the Kalman filter just before the filtering step. By replacing the SOC data in the Kalman filter with the SOC data in the microcomputer, It is possible to reflect the current during a short circuit.

[0160] In the SOC estimation process that performs the above calculations, a program that can execute the above formula is By transferring the data to a computer or microprocessor, the SOC can be calculated. It is also possible.

[0161] This embodiment can be combined with other embodiments. EXAMPLES

[0162] FIG. 7 is an example of a functional block for performing the calculation of the Kalman filter. The difference voltage indicated by reference numeral 401 in FIG. This is important for the occurrence of short circuits, and by monitoring this value, abnormalities in the secondary battery can be detected. .

[0163] A simulation was performed using data that simulated the current that periodically generated micro-short circuits. Perform a session.

[0164] Figure 8 shows the simulation results, with the horizontal axis representing time and the vertical axis representing the differential voltage. The figure shows the difference between the actual value (voltage) and the voltage estimated using the prior state variables (difference voltage).

[0165] In Figure 8, there are data showing periodic overshoots on the positive and negative sides. The data shows that there is a micro short circuit. Also, there is an overshoot only on the negative side. The data shown here is for comparison purposes, with no micro shorts. The overshoot is only observed on the charge side because the voltage fluctuates when switching from charge to discharge. In Fig. 8, overshoot is observed only on the negative side in the comparative example. The point to be measured is -0.0213V, so if you want to set a value greater than this, say -0.03V, By setting a threshold value, it is possible to prevent the error from being detected as an abnormality.

[0166] In addition, Figure 9 is a processed version of the data in Figure 8, with the vertical axis divided into left and right for easier viewing. In addition, the data used for the verification of Figure 8 was data in which the waveform of the micro-short was periodically inserted. In Fig. 8, the data is generated 13 times, but in an actual secondary battery, the data is generated randomly. The magnitude of the peak may also change depending on the usage status (charging or discharging) of the secondary battery. If a micro-short circuit is detected in either case, the deterioration of the secondary battery will be accelerated, or If the device is not detected, it will be unusable, so it is useful to be able to notify the user of the detection. The cause of the micro-short circuit has not yet been identified, but it is believed to be due to the inclusion of metal powder during manufacturing. There is also a theory that abnormalities cannot be detected immediately after manufacturing, but abnormal parts may appear after repeated charging and discharging. (conductive parts) grow and form, which may cause a micro-short circuit. When a secondary battery that generates heat is charged and discharged, it deteriorates rapidly and may suddenly become unusable. Therefore, the method of the present invention that is capable of detecting micro-short circuits is useful.

[0167] In the Kalman filter, the input values ​​of the battery are input to the battery equivalent circuit model, and their The outputs are compared, and if there is a difference, the difference is multiplied by the Kalman gain and fed back, so that the error is The battery equivalent circuit model is then modified so that it is minimized. This process is repeated successively.

[0168] Note that the Kalman filter is a system that adjusts sequentially, so the errors near the beginning of Figures 8 and 9 The difference is negligible.

[0169] The differential voltage value indicated by reference numeral 401 in FIG. 7 is −0.0631V as the minimum value and +0.03 The maximum voltage is 24V. Also, the peak voltage near 0 at the point where the micro short circuit occurred The points were -0.0386V on the negative side and +0.0186V on the positive side. Therefore, To detect all micro-shorts, set the threshold to -0.0386 on the negative side. V, +0.0186V on the positive side, and these values ​​can be detected using a comparator or similar. Since the battery life varies depending on the secondary battery used, it is necessary to appropriately simulate the battery life in advance using the characteristics data of the secondary battery used. A simulation can be performed and the threshold value, etc. can be determined based on the results.

[0170] The simulation results shown in Figures 8 and 9 are based on the circuit diagram provided by Analog Devices. LTspice (Simulation program with h integrated circuit emphasis. do. [Explanation of symbols]

[0171] 1 comparator, 2 comparator, 81: positive electrode, 82: negative electrode, 83: solid electrolyte layer, 84: substrate, 85: Wiring electrode, 86: Wiring electrode, 87: Positive electrode active material, 88: Negative electrode active material, 300: Abnormality Monitoring unit, 301: battery, 302: battery controller, 303: motor controller controller, 304: motor, 305: gear, 306: DCDC circuit, 307: electric power 308: Heater, 309: Defogger, 310: DCDC circuit, 311: Battery 312: Inverter, 313: Audio, 314: Power window, 315: Pumps, 316: Tires, 400: Micro short models, 401: Codes, 402: Retards 403: comparator, 404: neural network section, 405: AND circuit, 4 07: multiplexer, 600: secondary battery, 601: positive electrode cap, 602: battery can, 6 03: Positive electrode terminal, 604: Positive electrode, 605: Separator, 606: Negative electrode, 607: Negative electrode terminal 608: insulating plate, 609: insulating plate, 611: PTC element, 612: safety valve mechanism, 613 : Conductive plate, 614: Conductive plate, 1400: Storage battery, 1402: Positive electrode, 1404: Negative electrode, 80 21: charging device, 8022: cable, 8024: secondary battery, 8400: automobile, 840 1: Headlight, 8406: Electric motor, 8500: Car, 8600: Scooter, 8601: Side mirror, 8602: Secondary battery, 8603: Turn signal light, 8604: Seat Under Storage

Claims

[Claim 1] a first detection means for detecting a voltage value of the secondary battery, which is a first observation value; a second detection means for detecting a current value of the secondary battery, which is a second observation value; A calculation unit that calculates an estimated voltage value using a regression model; A secondary battery abnormality detection device having a judgment unit that calculates the difference between the voltage value of the first observation value and an estimated voltage value obtained at a previous time, and judges that the secondary battery is abnormal if the difference exceeds a certain threshold range.

Images