Fault monitoring method for sensor of energy storage system

Abstract

The invention relates to the technical field of energy storage systems, in particular to an energy storage system sensor fault monitoring method, which comprises the following steps: acquiring health data of an energy storage system sensor at historical nodes; the method comprises the following steps: matching a current monitoring time node with a historical node, obtaining theoretical data of an energy storage system sensor at the current monitoring time node, collecting real-time data, determining abnormal data, and determining a fault category of the energy storage system sensor by using a preset classification algorithm. According to the invention, by obtaining the historical node health data and matching the historical node health data with the current monitoring time node, a more accurate theoretical data reference can be provided for the current monitoring time node, so that abnormal data can be more accurately determined when the current monitoring time node is compared with real-time data; through rapid matching of the current monitoring time node and the historical node and rapid acquisition of theoretical data, real-time data can be rapidly analyzed in the monitoring process, abnormal conditions can be found in time, and the real-time performance of energy storage system sensor fault identification is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of energy storage systems, and in particular to a method for monitoring sensor faults in energy storage systems. Background Art

[0002] With the continuous development of the energy field, energy storage systems are playing an increasingly important role in power supply, renewable energy integration, etc. The stable operation of energy storage systems is crucial to ensuring the reliable supply of energy, and sensors are key components in energy storage systems. Their accurate monitoring and normal operation are the basis for ensuring the safe and efficient operation of energy storage systems.

[0003] However, in actual applications, energy storage system sensors may fail due to various reasons, such as aging, environmental impact, electrical interference, etc. Once a sensor fails, it may cause the operating status of the energy storage system to be unable to be accurately monitored, thereby affecting the performance and safety of the entire system. Traditional sensor fault monitoring methods often have problems such as low accuracy and slow response speed, which are difficult to meet the requirements of modern energy storage systems for high reliability and real-time performance. Summary of the invention

[0004] (I) Purpose of the invention

[0005] The object of the present invention is to provide a method for monitoring sensor failure of an energy storage system which can improve monitoring accuracy and real-time performance.

[0006] (II) Technical solution

[0007] To solve the above problems, the present invention provides a method for monitoring sensor failure in an energy storage system, comprising:

[0008] Obtain health data of energy storage system sensors at historical nodes;

[0009] Match the current monitoring time node with the historical node to obtain the historical node with the highest matching value, where the historical node with the highest matching value is the matching node;

[0010] According to the matching node, the theoretical data of the energy storage system sensor at the current monitoring time node is obtained;

[0011] Collect real-time data from energy storage system sensors at the current monitoring time node;

[0012] Determine abnormal data according to the theoretical data and the real-time data;

[0013] Determine the fault category of the energy storage system sensor according to the abnormal data using a preset classification algorithm;

[0014] The historical nodes are time nodes that divide a preset time range according to a preset time precision.

[0015] In another aspect of the present invention, preferably, obtaining health data of energy storage system sensors at historical nodes includes:

[0016] Obtaining the operating status and environmental status of the energy storage system at the historical node;

[0017] According to the operating state and environmental state of the energy storage system, obtaining first data of the energy storage system sensor at the historical node;

[0018] The first data is preprocessed to obtain second data, where the second data is health data of the energy storage system sensor at the historical node.

[0019] In another aspect of the present invention, preferably, preprocessing the first data to obtain the second data includes filtering the first data using the following formula to obtain the second data:

[0020]

[0021] Among them, h(r) represents the second data, f(r) represents the first data, f(ru) represents the value of the first data at position ru, σ represents the standard deviation of the Gaussian distribution, and u is a variable representing the offset from the center position.

[0022] In another aspect of the present invention, preferably, matching the current monitoring time node with the historical node to obtain the historical node with the highest matching value includes:

[0023] Calculate the inner matching value and the outer matching value of the current monitoring time node and the historical node respectively, wherein the inner matching value represents the matching value of the current monitoring time node and the historical node on the operating state of the energy storage system, and the outer matching value represents the matching value of the current monitoring time node and the historical node on the environmental state;

[0024] According to the inner matching value and the outer matching value, a historical node with the highest matching value is obtained.

[0025] In another aspect of the present invention, preferably, the inner matching value is calculated using the following formula:

[0026]

[0027] Among them, θ1 represents the inner matching value, x1…x n The characteristic vector representing the operating status of the energy storage system at the current monitoring time node, y1…y n represents the characteristic vector of the historical node in the operating state of the energy storage system, I represents the characteristic function,

[0028] In another aspect of the present invention, preferably, the outer matching value is calculated using the following formula:

[0029]

[0030] Among them, θ2 represents the external matching value, p i represents the ith feature vector of the current monitoring time node in the environmental state, m represents the total number of feature vectors, and q i represents the i-th feature vector of the historical node in the environment state, Represents the mean value of the environmental state feature vector at the current monitoring time node, Represents the mean of the feature vector of the historical node in the environment state.

[0031] In another aspect of the present invention, preferably, obtaining the historical node with the highest matching value according to the inner matching value and the outer matching value comprises calculating the matching value using the following formula:

[0032]

[0033] Among them, θ represents the matching value, θ1 represents the internal matching value, θ2 represents the external matching value, α represents the weight coefficient of the internal matching value, 1-α represents the weight coefficient of the external matching value, β1 represents the exponential coefficient of the internal matching value, and β2 represents the exponential coefficient of the external matching value.

[0034] In another aspect of the present invention, preferably, determining the fault category of the energy storage system sensor according to the abnormal data using a preset classification algorithm includes:

[0035] Preset the fault category of each type of sensor, and set corresponding characteristics for each fault category;

[0036] Determining the sensor type corresponding to the abnormal data;

[0037] Using a preset classification algorithm, calculate the posterior probability of the features corresponding to each fault category of the abnormal data and the corresponding sensor type;

[0038] The fault type of the energy storage system sensor is confirmed according to the posterior probability.

[0039] In another aspect of the present invention, preferably, a preset classification algorithm is used to determine the fault category of the energy storage system sensor, including using the following formula to confirm the fault category of the energy storage system sensor;

[0040]

[0041] Where C represents the fault category, F1,..F i ,..Fn represents the corresponding feature set of fault category C, P(C|F1,..F i ,..F n ) represents the posterior probability of fault category C of the sensor based on the corresponding feature set of fault category C, P(C) represents the prior probability of fault category C; P(F i |C) is the feature F in fault category C i Conditional probability of occurrence;

[0042] The class for which the logarithmic form of the posterior probability is maximized is the class of sensor failure.

[0043] In another aspect of the present invention, preferably, determining abnormal data according to the theoretical data and the real-time data comprises:

[0044] Calculating the difference between the theoretical data and the real-time data;

[0045] If the difference is greater than a preset threshold, the real-time data is abnormal data.

[0046] (III) Beneficial effects

[0047] The above technical solution of the present invention has the following beneficial technical effects:

[0048] The present invention can provide a more accurate theoretical data reference for the current monitoring time node by acquiring the health data of the historical nodes and matching them with the current monitoring time node, so as to more accurately determine the abnormal data when compared with the real-time data; the rapid matching of the current monitoring time node with the historical node and the rapid acquisition of theoretical data enable the real-time data to be analyzed quickly during the monitoring process, abnormal situations to be discovered in time, and the preset classification algorithm to determine the fault category of the energy storage system sensor, thereby improving the real-time performance of fault identification of the energy storage system sensor. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is an overall flow chart of an embodiment of the present invention. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical scheme and advantages of the present invention clearer, the present invention is further described in detail below in conjunction with specific embodiments and with reference to the accompanying drawings. It should be understood that these descriptions are only exemplary and are not intended to limit the scope of the present invention. In addition, in the following description, the description of well-known structures and technologies is omitted to avoid unnecessary confusion of the concept of the present invention.

[0051] Obviously, the described embodiments are only some embodiments of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] In the description of the present invention, it should be noted that the terms “first”, “second” and “third” are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0053] In addition, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0054] The present invention will be described in more detail below with reference to the accompanying drawings. In each of the accompanying drawings, the same elements are represented by similar reference numerals. For the sake of clarity, the various parts in the accompanying drawings are not drawn to scale.

[0055] Embodiment 1

[0056] A method for monitoring sensor failure in an energy storage system, Figure 1 FIG. 1 shows an overall flow chart of an embodiment of the present invention, as shown in FIG. Figure 1 As shown, including:

[0057] Obtain the health data of the energy storage system sensor of the historical node; extract the sensor data of the past period of time (such as one year, half a year or one quarter) from the database. The data should cover the sensor performance under various working conditions, including normal, minor abnormalities and serious failures. In this embodiment, according to the historical node, obtaining the health data of the energy storage system sensor of the historical node includes:

[0058] Determine the definition of historical nodes, which are time nodes divided by preset time range according to preset time precision. In the definition of historical nodes, the time range and precision need to be clarified. For example, in this embodiment, the past six months are selected as the time range, and the data is required to be accurate to the hour. It helps to obtain sufficiently rich and detailed historical data, providing a solid foundation for subsequent analysis. By defining such historical nodes, it can be ensured that the acquired data can fully reflect the performance of the energy storage system in different time periods and different working conditions.

[0059] Obtain the operating status and environmental status of the energy storage system at the historical node; after determining the historical node, further obtain the operating status and environmental status of the energy storage system corresponding to the historical node. The operating status includes the charging and discharging status, power output, current and voltage of the energy storage system. The environmental status includes external conditions such as temperature, humidity, and air pressure. The operating status and environmental status of the energy storage system directly affect the measurement value of the sensor.

[0060] According to the operating state and environmental state of the energy storage system, the first data of the energy storage system sensor at the historical node is obtained; after the operating state and environmental state of the energy storage system are obtained, the first data of the energy storage system sensor at the historical node is extracted according to the operating state and environmental state of the energy storage system. The first data includes the original value measured by the sensor, such as the temperature reading of the temperature sensor, the pressure reading of the pressure sensor, etc.

[0061] The first data is preprocessed to obtain second data, which is the health data of the energy storage system sensor at the historical node. Since the first data may be interfered by various factors, such as noise, drift, etc., the first data is preprocessed to obtain more accurate and reliable second data. The preprocessing process may include steps such as data cleaning (removing outliers, filling missing values, etc.), data smoothing (reducing noise), and data normalization (making data of different dimensions comparable). The second data after preprocessing is the health data of the energy storage system sensor at the historical node, because these data are closer to the actual performance of the sensor under ideal conditions.

[0062] In this embodiment, preprocessing the first data to obtain the second data includes filtering the first data using the following formula to obtain the second data:

[0063]

[0064] Among them, h(r) represents the second data, f(r) represents the first data, f(ru) represents the value of the first data at position ru, σ represents the standard deviation of the Gaussian distribution, and u is a variable representing the offset from the center position. Gaussian filtering can effectively smooth data and reduce the impact of noise. Compared with other filtering methods, Gaussian filtering can better retain the detailed information of the data while reducing noise.

[0065] Get the current monitoring time node, match the current monitoring time node with the historical node, and obtain the historical node with the highest matching value. The historical node with the highest matching value is the matching node; collect the sensor data of the current time node in real time as the current monitoring time node, match the current monitoring time node with the historical node, and calculate the matching degree. The matching degree can be based on a variety of factors, such as time, working conditions, sensor type, etc. Select the historical node with the highest matching value as the matching node, which will serve as the basis for subsequent analysis.

[0066] In this embodiment, the current monitoring time node and the historical node are matched to obtain the historical node with the highest matching value, and the historical node with the highest matching value is the matching node including:

[0067] Determine the sensor type: Before matching the current monitoring time node with the historical node, the sensor type must be determined. Different types of sensors may monitor different physical or chemical quantities, such as temperature, pressure, humidity, current, voltage, etc. Clarifying the sensor type helps to clarify the association between the data of the current monitoring time node and the data of the historical node, so as to perform accurate matching.

[0068] Calculate the internal matching value and external matching value of the current monitoring time node and the historical node respectively. The internal matching value indicates the matching value of the current monitoring time node and the historical node in the operating state of the energy storage system, and the external matching value indicates the matching value of the current monitoring time node and the historical node in the environmental state. The internal matching value calculates the matching degree of the current monitoring time node and the historical node in the operating state of the energy storage system. The operating state of the energy storage system involves the parameters of the energy storage system, such as charging and discharging state, power output, current and voltage, etc. The external matching value focuses on the matching degree of the current monitoring time node and the historical node in the environmental state. The environmental state may include external conditions such as temperature, humidity, and air pressure.

[0069] According to the internal matching value and the external matching value, the historical node with the highest matching value is obtained. After calculating the internal matching value and the external matching value respectively, the two values ​​can be weighted summed or otherwise combined to obtain a comprehensive matching value. This comprehensive matching value reflects the overall matching degree between the current monitoring time node and the historical node in terms of the operating status and environmental status of the energy storage system.

[0070] In this embodiment, the inner matching value is calculated using the following formula:

[0071]

[0072] Among them, θ1 represents the inner matching value, x1…x n The characteristic vector representing the operating status of the energy storage system at the current monitoring time node, y1…y n represents the characteristic vector of the historical node in the operating state of the energy storage system, I represents the indicative function, In this embodiment, the minimum and maximum ratio of the characteristic function is used to expand the distinction between the internal matching values ​​calculated between different historical nodes and the current monitoring time node in the running state. For example, the internal matching values ​​of the characteristic vectors of the two historical nodes and the current monitoring time node are 9.0005 and 9.00045 through the original calculation method, and the distinction is not obvious. Through the calculation method of this embodiment, the internal matching values ​​of the two historical nodes and the current monitoring time node are 8.13 and 8.2, and the distinction is obvious.

[0073] The outer match value is calculated using the following formula:

[0074]

[0075] Among them, θ2 represents the external matching value, p i represents the ith feature vector of the current monitoring time node in the environmental state, m represents the total number of feature vectors, and q i represents the i-th feature vector of the historical node in the environment state, Represents the mean value of the environmental state feature vector at the current monitoring time node, Represents the mean of the feature vector of the historical node in the environment state.

[0076] According to the inner matching value and the outer matching value, obtaining the historical node with the highest matching value includes calculating the matching value using the following formula:

[0077]

[0078] Among them, θ represents the matching value, θ1 represents the internal matching value, θ2 represents the external matching value, α represents the weight coefficient of the internal matching value, 1-α represents the weight coefficient of the external matching value, β1 represents the exponential coefficient of the internal matching value, and β2 represents the exponential coefficient of the external matching value. By comprehensively considering the operating status and environmental status of the energy storage system, and using the internal matching value and the external matching value to calculate the comprehensive matching value, the historical node that matches the current monitoring time node can be found more accurately.

[0079] According to the matching node, the theoretical data of the energy storage system sensor at the current monitoring time node is obtained; using the historical data of the matching node, through interpolation, fitting and other methods, the theoretical data of the sensor at the current monitoring time node is predicted;

[0080] Collecting real-time data of the energy storage system sensor at the current monitoring time node;

[0081] Determining abnormal data according to the theoretical data and the real-time data; including: calculating the difference between the theoretical data and the real-time data; if the difference is greater than a preset threshold, the real-time data is abnormal data.

[0082] According to the abnormal data, the fault category of the energy storage system sensor is determined using a preset classification algorithm; the preset classification algorithm, such as support vector machine (SVM), decision tree, neural network, etc., classifies the abnormal data. The classification algorithm can accurately identify different types of sensor faults, such as drift, failure, short circuit, etc. According to the classification result, the fault category of the sensor is determined.

[0083] In this embodiment, according to the abnormal data, using a preset classification algorithm, determining the fault category of the energy storage system sensor includes:

[0084] The fault category of each type of sensor is preset, and each fault category is set with corresponding features; the fault category of each type of sensor is preset according to the type of energy storage system sensor and the possible fault conditions. The fault category of each type of sensor is defined based on practical experience and professional knowledge to ensure that the possible fault conditions can be fully covered; for each fault category, the corresponding features are set. Features can be sensor readings, system status parameters, environmental parameters, etc., which can reflect the uniqueness and distinguishability of the fault category.

[0085] Determining the sensor type corresponding to the abnormal data;

[0086] Using a preset classification algorithm, calculate the posterior probability of the features corresponding to each fault category of the abnormal data and the corresponding sensor type;

[0087] The fault type of the energy storage system sensor is confirmed according to the posterior probability.

[0088] In this embodiment, the preset classification algorithm uses the following formula to confirm the fault category of the energy storage system sensor;

[0089]

[0090] Among them, C represents the category, F1,..F i ,..F n represents the corresponding feature set of category C, P(C|F1,..F i ,..F n ) represents the posterior probability that the fault of the sensor is of category C based on the corresponding feature set of category C, P(C) represents the prior probability of category C; P(F i |C) is the feature F in category C i The conditional probability of occurrence; the category that maximizes the logarithmic form of the posterior probability is the category of sensor failure.

[0091] The method further includes: issuing an early warning according to the fault category. In this embodiment, issuing an early warning according to the fault category includes:

[0092] Establish early warning notification channels; formulate corresponding early warning strategies according to the severity and urgency of the fault category. Establish early warning notification channels including sending early warning information to relevant personnel through sound and light alarms, SMS notifications, email reminders, etc.

[0093] Develop corresponding emergency response plans for different fault types; the plans should include steps such as fault identification, preliminary processing, emergency response, troubleshooting and repair to ensure that faults can be handled quickly and effectively when they occur.

[0094] Classify, store and manage the warning information. The classification basis can include fault type, occurrence time, processing results, etc. Through classified storage, the required warning information can be quickly located to improve processing efficiency.

[0095] The present invention can provide a more accurate theoretical data reference for the current monitoring time node by acquiring health data of historical nodes and matching them with the current monitoring time node, thereby more accurately determining abnormal data when compared with real-time data; the rapid matching of the current monitoring time node with the historical node and the rapid acquisition of theoretical data enable the real-time data to be analyzed quickly during the monitoring process and abnormal situations to be discovered in a timely manner.

[0096] It should be understood that the above specific embodiments of the present invention are only used to illustrate or explain the principles of the present invention, and do not constitute a limitation of the present invention. Therefore, any modifications, equivalent substitutions, improvements, etc. made without departing from the spirit and scope of the present invention should be included in the protection scope of the present invention. In addition, the appended claims of the present invention are intended to cover all changes and modifications that fall within the scope and boundaries of the appended claims, or the equivalent forms of such scope and boundaries.

[0097] The present invention has been described above with reference to the embodiments of the present invention. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of the present invention. The scope of the present invention is defined by the appended claims and their equivalents. Without departing from the scope of the present invention, a person skilled in the art may make various substitutions and modifications, which should all fall within the scope of the present invention.

[0098] Although the embodiments of the present invention have been described in detail, it should be understood that the various changes, substitutions, and alterations could be made hereto without departing from the spirit and scope of the invention.

[0099] Obviously, the above embodiments are merely examples for the purpose of clear explanation, and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived therefrom are still within the scope of protection of the invention.

Claims

1. A method for monitoring sensor failure in an energy storage system, characterized in that: include: Obtain health data of energy storage system sensors at historical nodes; Match the current monitoring time node with the historical node to obtain the historical node with the highest matching value, where the historical node with the highest matching value is the matching node; According to the matching node, the theoretical data of the energy storage system sensor at the current monitoring time node is obtained; Collect real-time data from energy storage system sensors at the current monitoring time node; Determine abnormal data according to the theoretical data and the real-time data; Determine the fault category of the energy storage system sensor according to the abnormal data using a preset classification algorithm; The historical nodes are time nodes that divide a preset time range according to a preset time precision.

2. The energy storage system sensor fault monitoring method according to claim 1, characterized in that: Obtaining the health data of energy storage system sensors at historical nodes includes: Obtaining the operating status and environmental status of the energy storage system at the historical node; According to the operating state and environmental state of the energy storage system, obtaining first data of the energy storage system sensor at the historical node; The first data is preprocessed to obtain second data, where the second data is health data of the energy storage system sensor at the historical node.

3. The energy storage system sensor fault monitoring method according to claim 2, characterized in that: Preprocessing the first data to obtain second data includes filtering the first data using the following formula to obtain second data: Among them, h(r) represents the second data, f(r) represents the first data, f(ru) represents the value of the first data at position ru, σ represents the standard deviation of the Gaussian distribution, and u is a variable representing the offset from the center position.

4. The energy storage system sensor fault monitoring method according to claim 2, characterized in that: Match the current monitoring time node with the historical node, and obtain the historical nodes with the highest matching value, including: Calculate the inner matching value and the outer matching value of the current monitoring time node and the historical node respectively, wherein the inner matching value represents the matching value of the current monitoring time node and the historical node on the operating state of the energy storage system, and the outer matching value represents the matching value of the current monitoring time node and the historical node on the environmental state; According to the inner matching value and the outer matching value, a historical node with the highest matching value is obtained.

5. The energy storage system sensor fault monitoring method according to claim 4, characterized in that: The inner match value is calculated using the following formula: Among them, θ1 represents the inner matching value, x1…x n The characteristic vector representing the operating status of the energy storage system at the current monitoring time node, y1…y n represents the characteristic vector of the historical node in the operating state of the energy storage system, I represents the characteristic function, 6. The energy storage system sensor fault monitoring method according to claim 5, characterized in that: The outer match value is calculated using the following formula: Among them, θ2 represents the external matching value, p i represents the ith feature vector of the current monitoring time node in the environmental state, m represents the total number of feature vectors, and q i represents the i-th feature vector of the historical node in the environment state, Represents the mean value of the environmental state feature vector at the current monitoring time node, Represents the mean of the feature vector of the historical node in the environment state.

7. The energy storage system sensor fault monitoring method according to claim 6, characterized in that: According to the inner matching value and the outer matching value, obtaining the historical node with the highest matching value includes calculating the matching value using the following formula: Among them, θ represents the matching value, θ1 represents the internal matching value, θ2 represents the external matching value, α represents the weight coefficient of the internal matching value, 1-α represents the weight coefficient of the external matching value, β1 represents the exponential coefficient of the internal matching value, and β2 represents the exponential coefficient of the external matching value.

8. The energy storage system sensor fault monitoring method according to claim 1, characterized in that: According to the abnormal data, using a preset classification algorithm, determining the fault category of the energy storage system sensor includes: Preset the fault category of each type of sensor, and set corresponding characteristics for each fault category; Determining the sensor type corresponding to the abnormal data; Using a preset classification algorithm, calculate the posterior probability of the features corresponding to each fault category of the abnormal data and the corresponding sensor type; The fault type of the energy storage system sensor is confirmed according to the posterior probability.

9. The energy storage system sensor fault monitoring method according to claim 8, characterized in that: Using a preset classification algorithm, determining the fault category of the energy storage system sensor, including using the following formula to confirm the fault category of the energy storage system sensor; Where C represents the fault category, F1,..F i ,..F n represents the corresponding feature set of fault category C, P(C|F1,..F i ,..F n ) represents the posterior probability of fault category C of the sensor based on the corresponding feature set of fault category C, P(C) represents the prior probability of fault category C; P(F i |C) is the feature F in fault category C i Conditional probability of occurrence; The class for which the logarithmic form of the posterior probability is maximized is the class of sensor failure.

10. The energy storage system sensor fault monitoring method according to claim 8, characterized in that: The determining of abnormal data according to the theoretical data and the real-time data comprises: Calculating the difference between the theoretical data and the real-time data; If the difference is greater than a preset threshold, the real-time data is abnormal data.

Images