Lithium battery thermal runaway monitoring device and method based on power line communication technology

Abstract

According to the lithium battery thermal runaway monitoring device and method based on the power line communication technology, early warning and accurate prevention and control of lithium battery thermal runaway are achieved through the innovative multi-sensor fusion monitoring network, the power line carrier communication and the artificial intelligence analysis technology. The system integrates a gas sensitive sensor, a distributed temperature sensor, a deformation sensor and an electrical parameter detection module, and constructs an omnibearing security situation awareness network; the power line carrier communication technology is adopted, efficient and reliable transmission of monitoring data is achieved in a modulation-demodulation mode, and system wiring is simplified; an intelligent analysis model is established based on a deep learning algorithm, time sequence data are processed through a long-short-term memory network, and a three-level early warning mechanism including an incubation period, an outbreak period and a spread period is constructed in combination with a fuzzy reasoning algorithm. The technical scheme has the remarkable advantages of convenient deployment, accurate monitoring, quick response, strong expansibility and the like, effectively solves the problems of complicated wiring, high response delay, high false alarm rate and the like of a traditional monitoring scheme, is particularly suitable for application scenes such as blade batteries and large-scale energy storage power stations, provides a brand new technical solution for safety monitoring of lithium batteries, and has a wide application prospect. And the method has important engineering application value.

Description

Technical Field

[0001] The present invention relates to the field of lithium battery safety monitoring technology, and specifically to a lithium battery thermal runaway monitoring device and method based on power line communication technology, which is particularly suitable for blade-type lithium battery packs and large-scale energy storage power station scenarios with limited space. Background Art

[0002] Current lithium battery thermal runaway monitoring technology has the following key defects:

[0003] (1) Communication architecture redundancy: Traditional solutions require independent deployment of dedicated communication cables (such as CAN bus or RS485 network), which leads to complex wiring and difficult maintenance in scenarios where battery modules are densely arranged, and the cable interfaces are susceptible to corrosion by electrolytes;

[0004] (2) Single monitoring dimension: Monitoring methods that rely on a single parameter (such as temperature or voltage) are unable to capture the co-evolution characteristics of multiple physical fields in the early stages of thermal runaway, resulting in a high false alarm rate;

[0005] (3) Significant response delay: The link delay from sensor data collection to central processor warning exceeds 3 seconds, while thermal runaway takes only 8-10 seconds from gas release to explosion. Existing technology cannot meet the millisecond-level response requirements; Summary of the Invention

[0006] 1. Purpose of the Invention

[0007] To solve the above problems, the present invention proposes a monitoring system combining multimodal perception, data fusion transmission, and artificial intelligence hierarchical response, aiming to:

[0008] (1) Eliminate the need for dedicated communication cable deployment by reusing power supply lines or wireless transmission channels;

[0009] (2) Integrate gas / temperature / shape / electricity four-dimensional sensor data to construct multi-parameter coupling criteria for early thermal runaway;

[0010] (3) Achieve millisecond-level early warning of thermal runaway processes.

[0011] 2. Technical Solution

[0012] 1. Device Architecture

[0013] like Figure 1 As shown, the system includes:

[0014] (a). Multi-parameter sensor group:

[0015] (1) Gas sensing unit: detects hydrogen, methane and volatile organic compounds (VOCs) in electrolyte, with detection limit as low as ppm level;

[0016] (2) Distributed temperature sensing network: monitors the temperature gradient between the battery surface and core area;

[0017] (3) Micro-strain deformation unit: captures the expansion deformation characteristics of the battery shell;

[0018] (4). Electrical parameter module: real-time collection of charging and discharging current and terminal voltage fluctuations.

[0019] (b) Edge computing unit: performs filtering, noise reduction, and feature extraction (such as temperature rise rate dT / dt and deformation acceleration) on the raw data;

[0020] (c) Adaptive communication module: supports power line carrier or low-power wireless transmission protocol;

[0021] (d) AI center: Contains a time series analysis model and a fuzzy inference engine, and outputs a three-level linkage warning signal.

[0022] 2. Core innovations

[0023] (a) Multi-source feature fusion mechanism:

[0024] A long short-term memory (LSTM) network is used to analyze four-dimensional time series data: hydrogen concentration curve, temperature gradient (shell surface vs. cell center), deformation accumulation, and voltage-SOC coupling matrix. The importance of features is dynamically weighted through an attention mechanism.

[0025] (b) Three-level warning logic:

[0026] Level 1 warning (latent period): The temperature rise rate continues to be ≥ 2°C / s, triggering the active cooling system;

[0027] Level 2 warning (outbreak period): When the hydrogen concentration is ≥500ppm and the core temperature is ≥60°C, the charging path is cut off and the audible and visual alarms are activated;

[0028] Level 3 warning (spreading period): deformation rate ≥ 5% accompanied by voltage fluctuation ≥ 20%, linkage fire extinguishing system and fault cluster isolation device. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 : System architecture block diagram (showing sensor → edge computing → communication → AI hub data flow);

[0030] Figure 2 : Three-level early warning linkage flow chart (including cooling / power off / fire extinguishing command trigger logic). DETAILED DESCRIPTION

[0031] Example 1: Vehicle-mounted blade battery pack monitoring

[0032] Take a blade battery pack with a length greater than 1m as an example:

[0033] 1. Sensor deployment optimization:

[0034] (1) The gas sensing unit is placed in the gas gathering area next to the module pressure relief valve, and the detection path length is ≤5cm;

[0035] (2) The temperature sensing chain adopts a distributed point strategy, with three temperature measurement nodes set at the positive electrode, negative electrode, and center of each battery;

[0036] (3) The deformation unit is attached to the stress concentration area on the long side of the battery.

[0037] 2. Anti-interference transmission design:

[0038] (1) The edge unit collects data in a 50ms cycle and transmits it via a battery DC power line carrier;

[0039] (2) The communication module uses frequency domain interleaving technology to suppress common mode interference.

[0040] 3. Thermal runaway interception verification:

[0041] When the LSTM model identifies that the dT / dt of a certain cell is continuously greater than 1.8°C / s (SOC ≥ 80%), the liquid cooling system performs targeted cooling to complete the cooling of the blade battery pack and suppress the occurrence of thermal runaway.

[0042] Example 2: 100MWh Energy Storage Power Station System

[0043] Take the containerized energy storage unit as an example:

[0044] 1. Layered processing architecture:

[0045] (1) Each battery cluster is equipped with an edge computing gateway and connected to the central server via a Mesh wireless network or power line carrier;

[0046] (2) The AI platform deploys a spatial clustering algorithm, combining the voltage differential matrix (ΔV_max / ΔV_min>2.0) with the gas concentration distribution heat map, and the positioning error is less than 1 battery module.

[0047] 2. Cross-system linkage:

[0048] When a level 3 warning is triggered, the coordinate information is synchronously pushed to the fire protection subsystem, with a full-link delay of less than 300ms; the warning log is automatically uploaded to the cloud-based operation and maintenance platform to support fault backtracking analysis.

[0049] Beneficial effects

[0050] Simplified deployment: Reusing power lines or wireless transmission significantly reduces wiring complexity and is suitable for high-density battery deployment scenarios.

[0051] False alarm rate optimization: Multi-sensor fusion reduces the false alarm rate to the leading level;

[0052] Breakthrough in response speed: The entire link delay from data collection to warning command output is less than 400ms, meeting the golden intervention window requirement for thermal runaway;

[0053] Improved scalability: Supports monitoring of energy storage stations at the thousand-node level, and the communication protocol is compatible with existing power infrastructure.

Claims

1. A lithium battery thermal runaway monitoring device based on power line communication technology, characterized in that: include: (a) Sensor group (1): used to monitor the status of lithium batteries in real time, including: A gas sensor (101) for detecting hydrogen, methane and electrolyte products released by thermal runaway of the lithium battery; A temperature sensor (102) detects temperature changes on the surface and inside of the lithium battery; A deformation sensor (103) monitors the deformation state of each wall surface of the lithium battery; A current and voltage detection module (104) collects lithium battery power supply current and voltage data in real time; (b) a front-end microcontroller (2), connected to the sensor group, for aggregating and pre-processing sensor data; (c) a signal modulation unit (3), which receives data from the front-end microcontroller and modulates the data into a power line carrier signal; (d) a power line communication network (4) for transmitting a modulated carrier signal; (e) a signal demodulation unit (5) for demodulating the carrier signal transmitted by the power line and restoring the sensing data; (f) A central processing unit (6) receives the demodulated data, analyzes the thermal runaway risk of the lithium battery through an artificial intelligence recognition method, and outputs the monitoring results.

2. The monitoring device according to claim 1, wherein: The gas sensor (101) is a multi-gas composite sensor, and its detection range covers hydrogen (0-10000ppm), methane (0-5000ppm) and electrolyte volatile organic compounds (VOCs); the temperature sensor (102) is a distributed optical fiber sensor or a thermocouple array, which is arranged at the positive electrode, negative electrode and center of the lithium battery; the deformation sensor (103) is a piezoelectric film or a strain gauge, which is attached to the surface of the lithium battery shell.

3. The monitoring device according to claim 1, wherein: The signal modulation unit (3) and the demodulation unit (5) adopt orthogonal frequency division multiplexing (OFDM) technology, with a carrier frequency range of 1-30 MHz and a transmission rate of ≥10 Mbps; the power line communication network (4) reuses the lithium battery power supply line, and no additional communication cables are required.

4. The monitoring device according to claim 1, wherein: The artificial intelligence recognition method of the central processing unit (6) includes: a time series data analysis model based on a long short-term memory network (LSTM), the input of which is time series data fused by multiple sensors; and the output of which is a thermal runaway risk level (level one warning: temperature anomaly; level two warning: gas concentration anomaly; level three warning: deformation / voltage anomaly accompanied by temperature / gas anomaly).

5. The monitoring device according to claim 1, wherein: The device is suitable for a blade-type lithium battery pack or a large-scale energy storage power station, and the sensor group is connected to the central processing unit (6) via a single line through a power line communication network (4).

6. A method for monitoring thermal runaway of a lithium battery based on power line communication technology, applied to the device according to any one of claims 1 to 5, characterized in that: The following steps are involved: Step S1: collecting hydrogen / methane concentration, temperature, deformation, and current and voltage data of the lithium battery in real time through a sensor group; Step S2: The front-end microcontroller (2) aggregates the data and performs filtering and standardization preprocessing; Step S3: the signal modulation unit (3) modulates the data into a power line carrier signal and transmits it through the power supply line; Step S4: The signal demodulation unit (5) demodulates the signal and uploads it to the central processing unit (6); Step S5: The central processing unit (6) analyzes the data through the AI model and triggers a graded warning if thermal runaway characteristics are detected: Level 1 warning: temperature rise rate ≥ 2°C / s; Level 2 warning: hydrogen concentration ≥500ppm and temperature ≥60℃; Level 3 warning: deformation ≥5% and accompanied by voltage fluctuation ≥20%.

7. The monitoring method according to claim 6, wherein: The AI model described in step S5 uses a fuzzy inference algorithm and combines the state of charge (SOC) and voltage differential matrix to evaluate the thermal runaway risk score.

8. The monitoring method according to claim 6, wherein: Real-time monitoring is started after the lithium battery is charged, and the early warning signal is synchronously transmitted to the terminal management platform through the power line communication network.

Images