Priority transmission method for abnormal triggering key data supporting QoS (Quality of Service)

Abstract

The invention discloses an abnormal trigger key data identification and priority transmission method supporting QoS, which is suitable for analyzing voltage waveform data output by a photovoltaic system, and comprises the following steps: 1, acquiring a voltage waveform data set output by the photovoltaic system, defining a real category label of each data point, and distinguishing normal waveform data from abnormal waveform data; 2, data decomposition: extracting each frequency component of the data layer by layer through a three-band filter, calculating the energy of each component, and predicting the abnormal probability of the data through LSTM; and 3, constructing a data backoff mechanism, and preferentially transmitting abnormal data. According to the invention, by combining the signal processing technology and the machine learning model, efficient analysis and anomaly detection of the voltage waveform data of the photovoltaic system are realized, and rapid transmission of key data is ensured, so that the performance and stability of the photovoltaic system are improved.

Description

Technical Field

[0001] The invention belongs to the technical field of data transmission, in particular to a method for preferentially transmitting key data triggered by abnormalities supporting QoS. Background Art

[0002] With the widespread application of photovoltaic power generation systems, real-time monitoring and efficient transmission of operating data have become the key to ensuring system stability and improving power generation efficiency. However, the voltage waveform data output by the photovoltaic system has waveform characteristics, which are usually manifested as sudden changes in voltage amplitude, and may cause instantaneous voltage drops or overvoltages in a short period of time. These signals may be caused by equipment failure, power system failure or external interference. Timely identification and priority transmission of these abnormal data are of great significance for the maintenance and fault warning of photovoltaic systems.

[0003] In the prior art, data transmission mostly adopts a unified priority scheduling strategy, lacking separate processing for abnormal data. This method can meet general needs under normal circumstances, but when the communication environment is complex or channel resources are limited, abnormal data may not be transmitted in time due to delay or loss, thus posing a threat to the fault detection and safe operation of the system. In addition, the existing anomaly detection methods usually rely on fixed thresholds or traditional statistical models, which are difficult to adapt to complex nonlinear time series data, resulting in low anomaly recognition rate and high misjudgment rate. On the other hand, with the increasing number of communication nodes in photovoltaic systems, the channel competition problem has become more and more significant. Most of the existing backoff mechanisms are not optimized for key data, and cannot guarantee the priority transmission of abnormal data under limited communication resources, further exacerbating the delay problem of key data. Therefore, the existing methods have obvious deficiencies in the detection and priority transmission of abnormal data, and there is an urgent need for an improved solution that can simultaneously improve the accuracy of abnormal recognition and transmission efficiency to meet the growing demand for data communication quality in photovoltaic systems. Summary of the invention

[0004] In view of the shortcomings of existing photovoltaic system monitoring and data transmission technologies, the present invention proposes a method for priority transmission of abnormally triggered key data that supports QoS, in order to achieve real-time and accurate monitoring and analysis of photovoltaic system voltage waveform data, thereby giving priority to key data during data transmission, thereby improving the operating stability and fault response speed of the photovoltaic system, reducing maintenance costs, and improving energy utilization efficiency.

[0005] In order to achieve the above-mentioned purpose, the present invention adopts the following technical scheme:

[0006] The method for preferentially transmitting critical data triggered by abnormalities supporting QoS of the present invention is characterized in that it is performed according to the following steps:

[0007] Step 1: Obtain the voltage waveform dataset output by the photovoltaic system ,in, represents the waveform data of the nth time period, where N is the total number of time periods; and , ) represents the waveform data of the tth time step in the nth period; T represents the total number of time steps in each period; definition for The true category label of 0,1}, if ,but For abnormal waveform data, if ,express It is normal waveform data;

[0008] Step 2: Define the number of decomposition layers as L, and After decomposition, low-frequency components, medium-frequency components, and high-frequency components are obtained, and then input into the LSTM model and output data The predicted probability of belonging to the anomaly class;

[0009] Step 3: According to the predicted probability that the waveform data belongs to the abnormal category, the waveform data of the abnormal category is divided, and a data backoff mechanism is constructed to give priority to the transmission of the divided abnormal data.

[0010] The method for identifying and preferentially transmitting abnormally triggered critical data supporting QoS described in the present invention is also characterized in that step 2 comprises:

[0011] Step 2.1: Define and initialize the current decomposition level ;Will As the j-1th low-frequency component of the nth time period ;

[0012] Step 2.2: Use a three-band filter to Perform the j-th layer extraction to obtain the j-th layer low-frequency component of the n-th time period , the j-th intermediate frequency component of the n-th time period , the j-th layer high-frequency component in the n-th period ,in, express The value of the intermediate frequency component at the kth time point; express The high-frequency component value at the kth time point;

[0013] Step 2.3: After assigning the value to j, return to step 2.2 and execute sequentially until So far, the waveform data after decomposition of the nth time period can be obtained by using formula (1): :

[0014] (1)

[0015] In formula (1), represents the Lth layer low-frequency component in the nth time period;

[0016] Step 2.4: Calculate using formula (2) Energy , and then use formula (3) to get Total Energy :

[0017] (2)

[0018] (3)

[0019] In formula (2), d is the weight of the low-frequency component, ;

[0020] Step 2.5: Use equation (4) and equation (5) to obtain The energy ratio and The total energy entropy :

[0021] (4)

[0022] (5)

[0023] Step 2.6: The total energy entropy Input into the LSTM model, and use formula (11) to get the tth time step The voltage characteristic of :

[0024] (6)

[0025] In formula (6), express The key feature of the voltage before the tth time step is obtained by formula (7); tanh represents the activation function, Indicates the tth time step The output gate of is obtained by formula (10);

[0026] (7)

[0027] In formula (7), express The key characteristics of the voltage before the t-1th time step, when t=1, are =0; Indicates the tth time step The forget gate is obtained by formula (8); is the tth time step The new voltage characteristic is expressed as follows and is obtained by equation (9);

[0028] (8)

[0029] In formula (8), represents the activation function, is the t-1th time step The voltage characteristic indicates that Represents the weight of the forget gate; represents the bias of the forget gate;

[0030] (9)

[0031] In formula (9), represents the weight of updating the state, Indicates the bias for updating the state;

[0032] (10)

[0033] In formula (10), Represents the weight of the output gate; represents the bias of the output gate;

[0034] Step 2.7: Using formula (11) we get Category score :

[0035] (11)

[0036] In formula (11), Represented as T time step is the feature representation of , W and b are the weight vector and bias of the last linear layer in the LSTM model respectively;

[0037] Step 2.8: Using formula (12) we get Predicted probability of belonging to the anomaly class :

[0038] (12)

[0039] Furthermore, the step 3 comprises:

[0040] Step 3.1: Build a cache queue for storing abnormal data and a cache queue for storing normal data ; Set threshold ,when When Store in abnormal data queue Otherwise, Store in normal data queue In; thus get and ,in, for The zth abnormal data in express The maximum length of for The vth normal data in the The maximum length of

[0041] Step 3.2: Define and initialize the number of exception backoffs =0 and normal backoff times =0, total abnormal backoff time = 0 and normal total backoff time , Maximum number of abnormal backoffs and the normal maximum backoff times , maximum total backoff time for exceptions and the normal maximum total backoff time ;

[0042] Step 3.3: If the buffer queue The first data in , the communication node scans the channel status. If the channel is occupied, no data is sent and step 3.6 is executed. If the channel is not occupied, data is sent directly. Then, proceed to step 3.4;

[0043] If the buffer queue The first data in , the communication node scans the channel status. If the channel is occupied, it will not send The first data in , and execute step 3.7. If the channel is not occupied, send Then, proceed to step 3.5;

[0044] Step 3.4: Assign to ,Will Assign to ,Will Assign to Then, return to step 3.3;

[0045] Step 3.5: Assign to ,Will Assign to ,Will Assign to Then, return to step 3.3;

[0046] Step 3.6: Assign to ,Will Assign to Then, execute step 3.8; is the abnormal single backoff time, and A random value between

[0047] Step 3.7: Assign to ,Will Assign to Then, execute step 3.9; is the normal single backoff time, and is a range of Random value;

[0048] Step 3.8: If = or = , then discard , and Assign to Then, return to step 3.3;

[0049] Step 3.9: If = or = , then discard , and Assign to Then return to step 3.3.

[0050] An electronic device of the present invention includes a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the priority transmission method, and the processor is configured to execute the program stored in the memory.

[0051] The present invention provides a computer-readable storage medium, wherein a computer program is stored on the computer-readable storage medium, and the computer program executes the steps of the priority transmission method when executed by a processor.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] 1. The present invention uses signal processing technology to perform multi-level decomposition of the voltage waveform data output by the photovoltaic system, combines the energy entropy characteristics with the time series modeling capability of the LSTM network, and realizes accurate identification of abnormal data.

[0054] 2. The present invention divides data into abnormal and normal categories, and combines data cache queues with dynamic backoff mechanisms to prioritize the transmission of abnormal data. This overcomes the problem in traditional data transmission methods where abnormal data and normal data have the same transmission priority, resulting in delay or loss of key data, and effectively reduces the average transmission delay of abnormal data. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 It is a structural diagram of the LSTM model of the present invention;

[0056] Figure 2 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0057] In this embodiment, a method for preferentially transmitting critical data triggered by abnormalities supporting QoS is provided, such as Figure 2 As shown, the steps are as follows:

[0058] Step 1: Obtain the voltage waveform dataset output by the photovoltaic system ,in, represents the waveform data of the nth time period, where N is the total number of time periods; and , ) represents the waveform data of the tth time step in the nth period; T represents the total number of time steps in each period; definition for The true category label of 0,1}, if ,but For abnormal waveform data, if ,express It is normal waveform data;

[0059] Step 2: Define the number of decomposition layers as L, and Decompose the waveform to obtain low-frequency components (usually representing the overall trend or long-term changes of the waveform, which may reflect the operating status of the system), medium-frequency components (including medium-duration characteristic information such as equipment failure or environmental changes), and high-frequency components (reflecting noise, spike signals, or instantaneous anomalies), which are input into the LSTM model and output data. The predicted probability of belonging to the anomaly class;

[0060] Step 2.1: Define and initialize the current decomposition level ;Will As the j-1th low-frequency component of the nth time period ;

[0061] Step 2.2: Use a three-band filter to Perform the j-th layer extraction to obtain the j-th layer low-frequency component of the n-th time period , the j-th intermediate frequency component of the n-th time period , the j-th layer high-frequency component in the n-th period ,in, express The value of the intermediate frequency component at the kth time point; express The high-frequency component value at the kth time point.

[0062] Step 2.3: After assigning the value to j, return to step 2.2 and execute sequentially until So far, the waveform data after decomposition of the nth time period can be obtained by using formula (1): :

[0063] (1)

[0064] In formula (1), represents the Lth layer low-frequency component in the nth time period;

[0065] Step 2.4: Calculate using formula (2) Energy , and then use formula (3) to get Total Energy :

[0066] (2)

[0067] (3)

[0068] In formula (2), d is the weight of the low-frequency component, .

[0069] Step 2.5: Use equation (4) and equation (5) to obtain The energy ratio and The total energy entropy :

[0070] (4)

[0071] (5)

[0072] Step 2.6: The total energy entropy Input into the LSTM model, and use formula (11) to get the tth time step The voltage characteristic of :

[0073] (6)

[0074] In formula (6), express The key feature of the voltage before the tth time step is obtained by formula (7); tanh represents the activation function, Indicates the tth time step The output gate of is obtained by formula (10);

[0075] (7)

[0076] In formula (7), express The key characteristics of the voltage before the t-1th time step, when t=1, are =0; Indicates the tth time step The forget gate is obtained by formula (8); is the tth time step The new voltage characteristic is expressed as follows and is obtained by equation (9);

[0077] (8)

[0078] In formula (8), represents the activation function, is the t-1th time step The voltage characteristic indicates that Represents the weight of the forget gate; represents the bias of the forget gate;

[0079] (9)

[0080] In formula (9), represents the weight of updating the state, Indicates the bias for updating the state;

[0081] (10)

[0082] In formula (10), Represents the weight of the output gate; represents the bias of the output gate;

[0083] Step 2.7: Using formula (11) we get Category score :

[0084] (11)

[0085] In formula (11), Represented as T time step is the feature representation of , W and b are the weight vector and bias of the last linear layer in the LSTM model respectively;

[0086] Step 2.8: Using formula (12) we get Predicted probability of belonging to the anomaly class :

[0087] (12).

[0088] Step 3: According to the predicted probability that the waveform data belongs to the abnormal category, the waveform data of the abnormal category is divided, and a data backoff mechanism is constructed to give priority to the transmission of the divided abnormal data;

[0089] Step 3.1: Build a cache queue for storing abnormal data and a cache queue for storing normal data , ensure the timely transmission of key data; set thresholds ,when When Store in abnormal data queue Otherwise, Store in normal data queue In; thus get and ,in, for The zth abnormal data in express The maximum length of for The vth normal data in the The maximum length of and V are both 6 packets (300B).

[0090] Step 3.2: Define and initialize the number of exception backoffs =0 and normal backoff times =0, total abnormal backoff time = 0 and normal total backoff time , Maximum number of abnormal backoffs and the normal maximum backoff times , , used to balance the number of retransmission attempts and communication efficiency, the maximum total backoff time for abnormalities and the normal maximum total backoff time ;

[0091] Step 3.3: Send abnormal data first, if the buffer queue The first data in , the communication node scans the channel status. If the channel is occupied, no data is sent and step 3.6 is executed. If the channel is not occupied, data is sent directly. Then, execute step 3.4; if the buffer queue The first data in , the communication node scans the channel status. If the channel is occupied, it will not send The first data in , and execute step 3.7. If the channel is not occupied, send Then proceed to step 3.5.

[0092] Step 3.4: Assign to ,Will Assign to ,Will Assign to Then, return to step 3.3;

[0093] Step 3.5: Assign to ,Will Assign to ,Will Assign to Then, return to step 3.3;

[0094] Step 3.6: Assign to ,Will Assign to Then, execute step 3.8; is the abnormal single backoff time, and A random value between

[0095] Step 3.7: Assign to ,Will Assign to Then, execute step 3.9; is the normal single backoff time, and is a range of Random value;

[0096] Step 3.8: If = or = , then discard , and Assign to Then, return to step 3.3;

[0097] Step 3.9: If = or = , then discard , and Assign to Then return to step 3.3.

[0098] In this embodiment, an electronic device includes a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the above method, and the processor is configured to execute the program stored in the memory.

[0099] In this embodiment, a computer-readable storage medium stores a computer program on the computer-readable storage medium, and the computer program executes the steps of the above method when executed by a processor.

Claims

1. A method for preferential transmission of critical data triggered by abnormalities supporting QoS, characterized in that: The steps are as follows: Step 1: Obtain the voltage waveform dataset output by the photovoltaic system ,in, represents the waveform data of the nth time period, where N is the total number of time periods; and , ) represents the waveform data of the tth time step in the nth period; T represents the total number of time steps in each period; definition for The true category label of 0,1}, if ,but For abnormal waveform data, if ,express It is normal waveform data; Step 2: Define the number of decomposition layers as L, and After decomposition, low-frequency components, medium-frequency components, and high-frequency components are obtained, and then input into the LSTM model and output data The predicted probability of belonging to the anomaly class; Step 3: According to the predicted probability that the waveform data belongs to the abnormal category, the waveform data of the abnormal category is divided, and a data backoff mechanism is constructed to give priority to the transmission of the divided abnormal data.

2. The method for identifying and prioritizing critical data triggered by abnormalities supporting QoS according to claim 2, characterized in that: The step 2 comprises: Step 2.1: Define and initialize the current decomposition level ;Will As the j-1th low-frequency component of the nth time period ; Step 2.2: Use a three-band filter to Perform the j-th layer extraction to obtain the j-th layer low-frequency component of the n-th time period , the j-th intermediate frequency component of the n-th time period , the j-th layer high-frequency component in the n-th period ,in, express The value of the intermediate frequency component at the kth time point; express The high-frequency component value at the kth time point; Step 2.3: After assigning the value to j, return to step 2.2 and execute sequentially until So far, the waveform data after decomposition of the nth time period can be obtained by using formula (1): : (1) In formula (1), represents the Lth layer low-frequency component in the nth time period; Step 2.4: Calculate using formula (2) Energy , and then use formula (3) to get Total Energy : (2) (3) In formula (2), d is the weight of the low-frequency component, ; Step 2.5: Use equation (4) and equation (5) to obtain The energy ratio and The total energy entropy : (4) (5) Step 2.6: The total energy entropy Input into the LSTM model, and use formula (11) to get the tth time step The voltage characteristic of : (6) In formula (6), express The key feature of the voltage before the tth time step is obtained by formula (7); tanh represents the activation function, Indicates the tth time step The output gate of is obtained by formula (10); (7) In formula (7), express The key characteristics of the voltage before the t-1th time step, when t=1, are =0; Indicates the tth time step The forget gate is obtained by formula (8); is the tth time step The new voltage characteristic is expressed as follows and is obtained by equation (9); (8) In formula (8), represents the activation function, is the t-1th time step The voltage characteristic indicates that Represents the weight of the forget gate; represents the bias of the forget gate; (9) In formula (9), represents the weight of updating the state, Indicates the bias for updating the state; (10) In formula (10), Represents the weight of the output gate; represents the bias of the output gate; Step 2.7: Using formula (11) we get Category score : (11) In formula (11), Represented as T time step is the feature representation of , W and b are the weight vector and bias of the last linear layer in the LSTM model respectively; Step 2.8: Using formula (12) we get Predicted probability of belonging to the anomaly class : (12)。 3. A method for identifying and prioritizing critical data triggered by abnormalities supporting QoS according to claim 3, characterized in that: The step 3 comprises: Step 3.1: Build a cache queue for storing abnormal data and a cache queue for storing normal data ; Set threshold ,when When Store in abnormal data queue Otherwise, Store in normal data queue In; thus get and ,in, for The zth abnormal data in express The maximum length of for The vth normal data in the The maximum length of Step 3.2: Define and initialize the number of exception backoffs =0 and normal backoff times =0, total abnormal backoff time = 0 and normal total backoff time , Maximum number of abnormal backoffs and the normal maximum backoff times , maximum total backoff time for exceptions and the normal maximum total backoff time ; Step 3.3: If the buffer queue The first data in , the communication node scans the channel status. If the channel is occupied, no data is sent and step 3.6 is executed. If the channel is not occupied, data is sent directly. Then, proceed to step 3.4; If the buffer queue The first data in , the communication node scans the channel status. If the channel is occupied, it will not send The first data in , and execute step 3.

7. If the channel is not occupied, send Then, proceed to step 3.5; Step 3.4: Assign to ,Will Assign to ,Will Assign to Then, return to step 3.3; Step 3.5: Assign to ,Will Assign to ,Will Assign to Then, return to step 3.3; Step 3.6: Assign to ,Will Assign to Then, execute step 3.8; is the abnormal single backoff time, and A random value between Step 3.7: Assign to ,Will Assign to Then, execute step 3.9; is the normal single backoff time, and is a range of Random value; Step 3.8: If = or = , then discard , and Assign to Then, return to step 3.3; Step 3.9: If = or = , then discard , and Assign to Then return to step 3.

3.

4. An electronic device, comprising a memory and a processor, characterized in that: The memory is used to store a program that supports a processor to execute the priority transmission method according to any one of claims 1 to 3, and the processor is configured to execute the program stored in the memory.

5. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the priority transmission method according to any one of claims 1 to 3 are executed.

Images