Virtual data calculation method of intelligent power transformation and distribution station house sensor

Abstract

The invention discloses a virtual data calculation method for a sensor of an intelligent power transformation and distribution station house, and the method comprises the steps: setting a real sensor data point or a virtual sensor data point in a sensor data point module, and enabling the real sensor data point to be a direct detection result of a sensor model; the virtual sensor data points are results obtained by performing operation on zero, one or more than two real sensor data points or other real sensor data points through an operation expression; in the working process of the sensor, a data change message of a real sensor data point on which a virtual sensor data point depends is subscribed, and when a change is detected, calculation of an operation expression is carried out in real time to obtain virtual sensor data. According to the method, by calculating the virtual sensor data points, real-time analysis and processing of the sensor data can be achieved, the efficiency and accuracy of sensor data processing can be improved, the excavation potential of existing sensor data is improved, and richer and more personalized data services are provided for application of the Internet of Things.

Description

Technical Field

[0001] The invention relates to a method for calculating virtual data of a sensor in a smart transformer substation, and belongs to a processing technology for virtual data of a sensor. Background Art

[0002] With the continuous development of science and technology, more and more sensors are needed in the abnormal detection process of substations. The analysis of sensor detection data is a very important part, which mainly includes the following aspects:

[0003] 1. Data collection and preprocessing: The sensor collects raw voltage, current, temperature and other operating data; performs preprocessing such as filtering and normalization on the collected raw data to improve data quality; and converts the format of the preprocessed data to facilitate subsequent analysis and calculation.

[0004] 2. Feature extraction and analysis: Extract relevant feature indicators from the original data according to the equipment operation characteristics, such as harmonic content, temperature rise, etc.; use statistical analysis, time-frequency analysis and other methods to conduct in-depth analysis of feature indicators and mine abnormal patterns; apply machine learning algorithms to cluster and associate feature data to discover potential fault patterns.

[0005] 3. Fault diagnosis and prediction: Establish a fault diagnosis model based on feature data to realize intelligent judgment of equipment fault type and severity; use time series analysis, deep learning and other methods to predict and model feature data to provide early warning of possible faults; integrate multi-source data, such as equipment operation logs, environmental data, etc., to improve the accuracy of fault diagnosis and prediction.

[0006] 4. Decision support and optimization: Combine fault diagnosis and prediction results with equipment maintenance strategies to provide decision support for operation and maintenance personnel; based on fault prediction results, use optimization algorithms to adjust equipment inspection and maintenance plans to improve resource utilization efficiency; use reinforcement learning and other technologies to continuously optimize fault diagnosis and prediction models and improve the intelligence level of the system.

[0007] In general, the analysis of sensor detection data is the core of distribution station anomaly detection. It needs to be combined with advanced signal processing, machine learning and other technologies to achieve intelligent monitoring and diagnosis of equipment operating status, and provide strong support for the safe and stable operation of the distribution system.

[0008] In order to achieve effective analysis of sensor detection data, the substation must first summarize and calculate the data of various sensors deployed in the substation, and then report it to the IoT platform through the power IoT technology. In actual operation, it is often necessary to combine multiple data points for data calculation and report the final data as reporting information. This data is not directly collected and reported by the sensor, but is obtained by the intelligent gateway after collecting one or more data points and using the designed algorithm for comprehensive calculation. We call such data sensor virtual data.

[0009] Virtual sensor data points are derived from the concept of data twins. Virtual sensor data points can be obtained by calculating real sensor data points. It is a method for real-time analysis and processing of real sensor data points. This method can not only ensure the real-time nature of data processing, but also improve the potential of data mining, thereby providing more personalized and accurate services to meet the data reporting needs of the IoT platform. Summary of the invention

[0010] Purpose of the invention: In order to overcome the deficiencies in the prior art, the present invention provides a virtual data calculation method for sensors in smart substations, which can not only ensure the real-time performance of data processing, but also improve the potential of data mining, thereby providing more personalized and accurate services to meet the data reporting requirements of the Internet of Things platform.

[0011] Technical solution: To achieve the above purpose, the technical solution adopted by the present invention is:

[0012] A method for calculating virtual data of a sensor in a smart substation comprises the following steps:

[0013] Step 1. In the gateway, a data model is constructed for each type of sensor, including a model recognition module and a model data point module. In the model data point module, real model data points or virtual model data points are set. Real model data points are the direct detection results of the data model itself, and virtual model data points are the results of operating zero, one, or more than two of itself or other real model data points through operation expressions; the operation expression is configured in the configuration of the virtual model data point, and the parsing method is selected as the operation expression parsing; the parsing method of the operation expression is written into the configuration dictionary of the gateway;

[0014] Step 2. In the gateway machine, a sensor model is created for each sensor, including a sensor identification module and a sensor data point module. Real sensor data points or virtual sensor data points are set in the sensor data point module. Real sensor data points are direct detection results of the sensor model itself, and virtual sensor data points are the results of computing zero, one, or more than two of itself or other real sensor data points through operation expressions. The sensor identification module selects a data model based on the sensor type, and calls the model identification module of the selected data model as the design basis. The sensor data point module selects a data model based on the sensor type, and calls the model data point module of the selected data model as the design basis. The operation expression involved in the virtual sensor data point calls the operation expression of the selected data model, and selects the real model data point as the real sensor data point.

[0015] Step 3: The gateway loads all sensor configurations and all sensor data point configurations to complete sensor initialization;

[0016] Step 4, traverse the sensor to determine whether it is a real sensor data point or a virtual sensor data point: if it is a real sensor data point, continue traversing; if it is a virtual sensor data point, find out the list of real sensor data points that the virtual sensor data point depends on according to the analytical method of the operation expression;

[0017] Step 5, store the virtual sensor data points and the list of dependent real sensor data points into the processing cache;

[0018] Step 6. Subscribe to the data change messages of the real sensor data points that it depends on;

[0019] Step 7: If the data of the real sensor data point that is relied on changes, the operation expression is started to calculate and obtain the virtual sensor data point.

[0020] Specifically, in the data model of the sensor, the model identification module and the model data point module are both designed as tables, respectively recorded as model table and model data point table; in the sensor model, the sensor identification module and the sensor data point module are both designed as tables, respectively recorded as sensor table and sensor data point table.

[0021] Specifically, the model table at least includes the field id representing the model ID, the field name representing the model name, and the field datasourcetype representing the model data source type. The data type of the field id is Int, the data type of the field name is String, and the data type of the field datasourcetype is String; the model data point table at least includes the field id representing the model data point ID, the field name representing the model data point name, the field modelid representing the model ID, the field remotetype representing the model remote type, and the field url representing a real model data point or a virtual model data point. The data type of the field id is Int, the data type of the field name is String, the data type of the field modelid is Int, the data type of the field remotetype is String, and the data type of the field url is String. If the field url represents a virtual model data point, the field url records the operation expression code of the virtual model data point; the value of the field modelid in the model data point table is equal to the field id in the model table.

[0022] Specifically, the parsing method of the operation expression is written into the configuration dictionary of the gateway machine; the configuration dictionary of the gateway machine is designed with an operation expression parsing list, and the operation expression parsing list at least includes an expression ID and an expression operation parsing code corresponding to the expression ID.

[0023] Specifically, the sensor table includes at least the field id representing the sensor ID, the field name representing the sensor name, and the field modelid representing the model ID. The data type of the field id is Int, the data type of the field name is String, and the data type of the field modelid is Int. The sensor data point table includes at least the field id representing the model data point ID, the field name representing the model data point name, the field sensorid representing the sensor ID, the field modelpointid representing the sensor data point ID, and the field url representing the real sensor data point or the virtual sensor data point. The data type of the field id is Int, the data type of the field name is String, and the data type of the field sensorid is Int, the data type of the field modelpointid is Int, the data type of the field url is String. If the field url represents a virtual sensor data point, the field url records the operation expression of the virtual sensor data point; the value of the field modelid in the sensor table is the field id in the model table; the value of the field id in the sensor data point table is equal to the field id in the model data point table, the value of the field name in the sensor data point table is equal to the field name in the model data point table, the value of the field sensorid in the sensor data point table is equal to the field id in the sensor table, and the model data point ID involved in the operation expression in the field url in the model data point table is replaced with the sensor data point ID as the value of the field url in the sensor data point table.

[0024] Specifically, the operation expression supports numbers and operation symbols, and the operation symbols include (,), +, -, *, / , &&, ||, !, %; the representation format of the real model data point or virtual model data point is M: model data point ID, and the representation format of the real sensor data point or virtual sensor data point is S: sensor data point ID; when editing the operation expression, different operation symbols or numbers are separated by spaces; after the operation expression is edited, first verify the operation expression, and save the configuration after the verification passes.

[0025] Specifically, in Step 3, when configuring the sensor data point, if the real sensor data point is a real-time variable, the initial value of the real sensor data point is set to an invalid value. When the real sensor data point acquisition fails, the initial value is used as the real sensor data point; when configuring the sensor data point, if the real sensor data point is a fixed value, the initial value of the real sensor data point is set to a specified fixed value.

[0026] Beneficial effect: The virtual data calculation method of the smart substation room sensor provided by the present invention can obtain virtual sensor data points through accurate collection, real-time subscription and linkage calculation technology of real sensor data points, and then realize real-time analysis and processing of sensor data, which can improve the efficiency and accuracy of sensor data processing, increase the potential for mining existing sensor data, and provide richer and more personalized data services for Internet of Things applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is an implementation flow chart of the present invention. DETAILED DESCRIPTION

[0028] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0029] like Figure 1 The figure shows a method for calculating virtual data of sensors in a smart substation, and each step is described in detail below in conjunction with an embodiment.

[0030] Step 1. Build a data model for each type of sensor in the gateway machine.

[0031] In order to avoid duplicate configuration, it is necessary to classify the sensors in the station house and build a data model for each type of sensor in the gateway. The data model is the basis for creating all sensors of the same type.

[0032] The data model includes a model identification module and a model data point module. Real model data points or virtual model data points are set in the model data point module. The real model data points are the direct detection results of the data model itself, and the virtual model data points are the results of operating zero, one or more of itself or other real model data points through operation expressions; the operation expression is configured to the configuration of the virtual model data point, and the parsing method is selected as the operation expression parsing; the parsing method of the operation expression is written into the configuration dictionary of the gateway machine.

[0033] In this example, the model identification module and the model data point module in the data model are both designed as tables, which are respectively recorded as the model table and the model data point table.

[0034] Table 1 is an example of a model table, including but not limited to the table content.

[0035] Table 1 Model table example

[0036] Fields type illustrate id Int Model ID, unique name String Model name, unique datasourcetype String Model data source type, whether it is wireless or wired, etc.

[0037] The model data point module is all the data information contained in the data model. When creating a sensor, all model data points can be obtained according to the selected data model, and then sensor data points can be created according to the relevant information of the model data points. Table 2 shows an example of a model data point table, including but not limited to the table content.

[0038] Table 2 Example of model data point table

[0039]

[0040] Note: The value of the modelid field in Table 2 is equal to the id field in Table 1.

[0041] There is no virtual sensor data point for the actual sensor. Therefore, when configuring the model data point, you need to add a virtual model data point. This data point is not the data point actually sampled by the sensor, but is obtained through comprehensive calculation of 0 (not including sensor data points, only mathematical expressions, such as fixed values), 1 or more sensor data points (calculated by operation expression). The data analysis method for this data point is operation expression analysis. Select the model data point that needs to participate in the operation according to the needs, and edit the calculation method as operation expression.

[0042] The operation expression supports numbers and operation symbols, including (,), +, -, *, / , &&, ||, !, %; the representation format of real model data points or virtual model data points is M: model data point ID, and different operation symbols or numbers are separated by spaces; after the operation expression is edited, first verify the operation expression, and save the configuration after the verification passes.

[0043] If you want to configure the temperature and humidity over-limit alarm of the wireless temperature and humidity sensor, the condition is: when the temperature is greater than 30°C and the humidity is greater than 75, the value of the alarm data point is 1; otherwise it is 0. Assuming that the temperature model data point ID of the wireless temperature and humidity model is 23345 and the humidity model data point ID is 23346, the configured expression is: (M:23345>30)&&(M:23346>75); the data type is bool.

[0044] The configuration dictionary of the gateway machine is designed with an operation expression parsing list, which includes at least an expression ID and an expression operation parsing code corresponding to the expression ID. Table 3 shows an example of an operation expression parsing list, including but not limited to the table content.

[0045] Table 3 Example of operation expression analysis list

[0046] Dictionary code Expression ID Operation expression parsing code datadecodetype 201 Operation expression analysis

[0047] Step 2. In the gateway machine, create a sensor model for each sensor.

[0048] In the gateway machine, a sensor model is created for each sensor, including a sensor identification module and a sensor data point module. Real sensor data points or virtual sensor data points are set in the sensor data point module. The real sensor data points are the direct detection results of the sensor model itself, and the virtual sensor data points are the results of operating zero, one or more than two of itself or other real sensor data points through operation expressions; the sensor identification module selects a data model based on the sensor type, and calls the model identification module of the selected data model as the design basis; the sensor data point module selects a data model based on the sensor type, and calls the model data point module of the selected data model as the design basis; the operation expression involved in the virtual sensor data point calls the operation expression of the selected data model, and selects the real model data point as the real sensor data point.

[0049] In this example, the sensor identification module and the sensor data point module in the sensor model are both designed as tables, which are respectively recorded as sensor table and sensor data point table. After the sensor is created, the sensor data is stored in the sensor table, and the sensor data point information is stored in the sensor data point table.

[0050] Table 4 shows an example of a sensor table, including but not limited to the table contents.

[0051] Table 4 Sensor table example

[0052] Fields type illustrate id Int Sensor ID, unique name String Sensor name, unique modelid Int Model ID

[0053] Note: The value of the modelid field in Table 4 is the same as the id field in Table 1.

[0054] Table 5 shows an example of a sensor data point table, including but not limited to the table content.

[0055] Table 5 Sensor data point table example

[0056]

[0057] Note: The value of the field id in Table 5 is equal to the field id in Table 2, the value of the field name in Table 5 is equal to the field name in Table 2, and the value of the field sensorid in Table 5 is equal to the field id in Table 4. Generally, the url of a sensor data point is inherited from the corresponding model data point, but it can also be modified according to the characteristics of the sensor.

[0058] When creating a sensor, select the corresponding data model, and the sensor data point will be created based on the model data point, including real sensor data point and virtual sensor data point. When creating the sensor data point, you need to do some processing here, translate the model data point in the operation expression into the sensor data point. The representation format of the real sensor data point or virtual sensor data point is S: sensor data point ID. Note that different operation symbols or numbers are separated by spaces. After the operation expression is edited, first verify the operation expression, and save the configuration after the verification passes. For example, the ID of the created wireless sensor temperature data point is 45782, and the ID of the humidity data point is 45783. When creating the sensor temperature and humidity over-limit alarm data point, change (M:23345>30)&&(M:23346>75) to (S:45782>30)&&(S:45783>75).

[0059] Step 3: The gateway loads all sensor configurations and all sensor data point configurations to complete sensor initialization.

[0060] Sometimes a sensor data point fails to be sampled. To deal with this situation, when configuring the sensor data point, if the real sensor data point is a real-time variable, the initial value of the real sensor data point is set to an invalid value (for example, -99999). When the real sensor data point fails to be collected, the initial value (that is, the invalid value -99999) is used as the real sensor data point to be brought into the operation expression for calculation. When the operation expression encounters an invalid value (for example, -99999), a comprehensive judgment is required: if the value needs to be used for calculations such as +, -, *, / , &&, the real sensor data point is determined to be invalid, and the final result of the virtual sensor data point is also an invalid value (for example, -99999); if the value is not needed for calculation, the value can be ignored and the operation expression can be directly calculated.

[0061] When configuring the sensor data point, if the real sensor data point is a fixed value and does not involve calculation of other real sensor data points, then when configuring the sensor data point, the initial value of the real sensor data point will be set to the specified fixed value.

[0062] Step 4: Traverse the sensors and determine whether they are real sensor data points or virtual sensor data points.

[0063] Traverse the sensors and determine whether they are real sensor data points or virtual sensor data points based on whether the parsing method is an operation expression: if they are real sensor data points, continue traversing; if they are virtual sensor data points, find the list of real sensor data points that the virtual sensor data points depend on based on the parsing method of the operation expression.

[0064] Step 5: Store the virtual sensor data points and the list of dependent real sensor data points into the processing cache.

[0065] Step 6: Subscribe to the data change messages of the real sensor data points that it depends on.

[0066] Step 7: If the data of the real sensor data point that is relied on changes, the operation expression is started to calculate and obtain the virtual sensor data point.

[0067] The process of starting the calculation expression is as follows:

[0068] First, replace S: sensor data point ID in the virtual sensor data point expression one by one with the actual value of the sensor, such as the temperature and humidity over-limit alarm data point expression: (S:45782>30)&&(S:45783>75); if the temperature value is 35 and the humidity value is 55 at this time, the replaced expression is: (35>30)&&(55>75).

[0069] Then, according to the rule of calculating the expression inside the brackets first and then the expression outside the brackets, the calculation results of each step are as follows:

[0070] (35>30)&&(55>75)

[0071] 1&&(55>75)

[0072] 1&&0 0

[0074] The final calculation result is 0, which is the final result of the temperature and humidity over-limit alarm data point.

[0075] By calculating virtual sensor data points in this way, virtual sensor data points can be calculated dynamically and timely, and data can be reported according to platform requirements.

[0076] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the above embodiments do not limit the present invention in any form, and any technical solution obtained by equivalent replacement or equivalent transformation falls within the protection scope of the present invention.

Claims

1. A virtual data calculation method for sensors in a smart substation, characterized in that: The steps include: Step 1. In the gateway machine, a data model is constructed for each type of sensor, including a model recognition module and a model data point module. In the model data point module, real model data points or virtual model data points are set. Real model data points are the direct detection results of the data model itself, and virtual model data points are the results of operating zero, one, or more than two of itself or other real model data points through operation expressions; the operation expression is configured in the configuration of the virtual model data point, and the parsing method is selected as the operation expression parsing; Step 2. In the gateway machine, a sensor model is created for each sensor, including a sensor identification module and a sensor data point module. Real sensor data points or virtual sensor data points are set in the sensor data point module. Real sensor data points are direct detection results of the sensor model itself, and virtual sensor data points are the results of computing zero, one, or more than two of itself or other real sensor data points through operation expressions. The sensor identification module selects a data model based on the sensor type, and calls the model identification module of the selected data model as the design basis. The sensor data point module selects a data model based on the sensor type, and calls the model data point module of the selected data model as the design basis. The operation expression involved in the virtual sensor data point calls the operation expression of the selected data model, and selects the real model data point as the real sensor data point. Step 3: The gateway loads all sensor configurations and all sensor data point configurations to complete sensor initialization; Step 4, traverse the sensors and determine whether it is a real sensor data point or a virtual sensor data point based on whether the parsing method is an operation expression: if it is a real sensor data point, continue traversing; if it is a virtual sensor data point, find out the list of real sensor data points that the virtual sensor data point depends on based on the parsing method of the operation expression; Step 5, store the virtual sensor data points and the list of dependent real sensor data points into the processing cache; Step 6. Subscribe to the data change messages of the real sensor data points that it depends on; Step 7: If the data of the real sensor data point that is relied on changes, the operation expression is started to calculate and obtain the virtual sensor data point.

2. The virtual data calculation method of the smart substation sensor according to claim 1 is characterized by: In the data model of the sensor, the model identification module and the model data point module are both designed as tables, respectively recorded as model table and model data point table; in the sensor model, the sensor identification module and the sensor data point module are both designed as tables, respectively recorded as sensor table and sensor data point table.

3. The virtual data calculation method of the smart substation sensor according to claim 2 is characterized by: The model table at least includes the field id representing the model ID, the field name representing the model name, and the field datasourcetype representing the model data source type. The data type of the field id is Int, the data type of the field name is String, and the data type of the field datasourcetype is String; the model data point table at least includes the field id representing the model data point ID, the field name representing the model data point name, the field modelid representing the model ID, the field remotetype representing the model remote type, and the field url representing a real model data point or a virtual model data point. The data type of the field id is Int, the data type of the field name is String, the data type of the field modelid is Int, the data type of the field remotetype is String, and the data type of the field url is String. If the field url represents a virtual model data point, the field url records the operation expression code of the virtual model data point; the value of the field modelid in the model data point table is equal to the field id in the model table.

4. The virtual data calculation method of the smart substation sensor according to claim 2 is characterized by: The analysis method of the operation expression is written into the configuration dictionary of the gateway machine; the configuration dictionary of the gateway machine is designed with an operation expression analysis list, and the operation expression analysis list at least includes an expression ID and an expression operation analysis code corresponding to the expression ID.

5. The virtual data calculation method of the smart substation sensor according to claim 2 is characterized by: The sensor table at least includes the field id representing the sensor ID, the field name representing the sensor name, and the field modelid representing the model ID. The data type of the field id is Int, the data type of the field name is String, and the data type of the field modelid is Int. The sensor data point table at least includes the field id representing the model data point ID, the field name representing the model data point name, the field sensorid representing the sensor ID, the field modelpointid representing the sensor data point ID, and the field url representing the real sensor data point or the virtual sensor data point. The data type of the field id is Int, the data type of the field name is String, and the data type of the field sensorid is Int. t, the data type of the field modelpointid is Int, the data type of the field url is String, if the field url represents a virtual sensor data point, then the field url records the operation expression of the virtual sensor data point; the value of the field modelid in the sensor table is the field id in the model table; the value of the field id in the sensor data point table is equal to the field id in the model data point table, the value of the field name in the sensor data point table is equal to the field name in the model data point table, the value of the field sensorid in the sensor data point table is equal to the field id in the sensor table, and the model data point ID involved in the operation expression in the field url in the model data point table is replaced with the sensor data point ID as the value of the field url in the sensor data point table.

6. The virtual data calculation method of the smart substation sensor according to claim 2 is characterized by: The operation expression supports numbers and operation symbols, including (,), +, -, *, / , &&, ||, !, %; the representation format of real model data points or virtual model data points is M: model data point ID, and the representation format of real sensor data points or virtual sensor data points is S: sensor data point ID; when editing the operation expression, different operation symbols or numbers are separated by spaces; After editing the operation expression, verify the operation expression first, and save the configuration after verification.

7. The virtual data calculation method of the smart substation sensor according to claim 1 is characterized by: In the Step 3, when configuring the sensor data point, if the real sensor data point is a real-time variable, the initial value of the real sensor data point is set to an invalid value. When the real sensor data point acquisition fails, the initial value is used as the real sensor data point. When configuring the sensor data point, if the real sensor data point is a fixed value, the initial value of the real sensor data point is set to a specified fixed value.

Images