Equipment state evaluation method, electronic equipment and system

Abstract

The invention provides an equipment state evaluation method, electronic equipment and an equipment state evaluation system. The method comprises the following steps: predicting the generated power of a prediction time period based on the generated power of historical power generation of a generator set; the predicted generated power is compared with the actual generated power of the generator set in the prediction time period, and whether fault analysis is carried out on the generator set or not is judged; if the generator set is subjected to fault analysis, respectively constructing a frequency analysis chart and a loudness analysis chart based on historical audio data of the generator set; constructing a fusion analysis graph based on the frequency analysis graph and the loudness analysis graph, inputting the fusion analysis graph into a pre-trained neural network model, and outputting a fault probability by the neural network model; and determining whether the equipment state is a fault state based on the fault probability. According to the scheme, firstly, the equipment state is preliminarily determined through the predicted generated power, and whether the equipment is in the fault state or not is determined through analysis of the audio data collected in the corresponding time period.

Description

Technical Field

[0001] The present invention relates to the technical field of power plant management, and in particular to an equipment status assessment method, electronic equipment and system. Background Art

[0002] Power plant equipment management plays a vital role in ensuring the stability and security of power supply. As the core link of power production, the operating status of power plant equipment directly affects the output quality of electric energy and the continuity of power supply. Therefore, scientific and efficient management of power plant equipment is the key to ensuring stable power supply, improving energy utilization efficiency, and reducing downtime. In the context of increasingly fierce market competition and changing energy policies, the importance of power plant equipment management is more prominent. By introducing advanced information and intelligent technologies, remote monitoring, intelligent early warning and fault diagnosis of equipment can be realized, and the efficiency and accuracy of equipment management can be further improved. This not only helps power plants cope with various emergencies, but also provides strong support for the sustainable development of power plants.

[0003] In summary, power plant equipment management is an important part of the power production process and is directly related to the stability and safety of power supply. Therefore, power plants should attach great importance to equipment management, continuously improve management systems, and enhance management levels to ensure the smooth progress of power production and the long-term stable development of power plants. However, the existing technology for equipment management often relies on manual inspections, which is inefficient.

[0004] In view of this, the present invention is proposed. Summary of the invention

[0005] The purpose of the present invention is to provide a device status assessment method, electronic device and system. This scheme analyzes the sound frequency and loudness generated during the operation of the device through audio data collected in the corresponding time period to determine whether the device is in a fault state.

[0006] The present invention provides a method for evaluating equipment status, the method comprising the following steps:

[0007] Predicting the power generation during the forecast period based on the historical power generation of the power generating set;

[0008] Compare the predicted power generation with the actual power generation of the generator set during the predicted time period to determine whether to perform a fault analysis on the generator set;

[0009] If a fault analysis is performed on the generator set, a frequency analysis graph and a loudness analysis graph are respectively constructed based on the historical audio data of the generator set;

[0010] Constructing a fusion analysis graph based on the frequency analysis graph and the loudness analysis graph, inputting the fusion analysis graph into a pre-trained neural network model, and the neural network model outputting a fault probability;

[0011] It is determined whether the device state is a fault state based on the fault probability.

[0012] By adopting the above scheme, this scheme can first automatically count the power generation, and preliminarily determine the equipment status by comparing the predicted power generation with the actual power generation of the generator set in the predicted time period, and determine whether the equipment needs to be judged for faults. It can further analyze the audio data collected in the corresponding time period through the sound frequency and loudness generated during the operation of the equipment to determine whether the equipment is in a fault state.

[0013] In some embodiments of the present invention, in the step of determining whether the device state is a fault state based on the fault probability, if the device state is determined to be a fault state, historical audio data of the corresponding time period is recorded and uploaded.

[0014] In some embodiments of the present invention, in the step of predicting the power generation in a prediction time period based on the power generation of the historical power generation of the generator set, the power generation at the same time point every day is constructed as a corresponding prediction vector, and the prediction vector is input into a preset power prediction model. The power prediction model outputs the power generation at the corresponding time point of the prediction day, and the predicted power generation at each time point in the prediction time period is determined in the same manner, and the average value of the power generation at each time point in the prediction time period is calculated as the power generation for the prediction time period.

[0015] In some embodiments of the present invention, in the step of comparing the predicted power generation with the actual power generation of the generator set in the prediction time period to determine whether to perform a fault analysis on the generator set, if the predicted power generation calculated is less than the actual power generation of the generator set in the prediction time period, and the difference between the predicted power generation and the actual power generation of the generator set in the prediction time period is less than a preset power threshold, it is determined to perform a fault analysis on the generator set, otherwise, there is no need to perform a fault analysis on the generator set.

[0016] In some embodiments of the present invention, in the steps of respectively constructing a frequency analysis graph and a loudness analysis graph based on the historical audio data of the generator set, a line graph with time as the horizontal axis and frequency value as the vertical axis is constructed based on the frequency data of the predicted time period; and a line graph with time as the horizontal axis and loudness value as the vertical axis is constructed based on the loudness data of the predicted time period.

[0017] In some embodiments of the present invention, the step of constructing a fusion analysis graph based on the frequency analysis graph and the loudness analysis graph includes:

[0018] Calculate the difference between the frequency value at each time point in the frequency analysis graph and the preset standard frequency value, and normalize it;

[0019] Calculate the difference between the loudness value at each time point in the loudness analysis graph and the preset standard loudness value, and normalize it;

[0020] For each time point, the weighted average of the difference between the normalized frequency value and the preset standard frequency value and the weighted average of the difference between the loudness value and the preset standard loudness value are calculated as the fusion value, and a line graph with time as the horizontal axis and the calculated fusion value as the vertical axis is constructed as a fusion analysis graph.

[0021] By adopting the above scheme, this scheme can construct feature analysis graphs corresponding to the two features respectively by statistically analyzing the frequency and loudness, and further combine the two features to construct a fusion analysis graph. In the fusion analysis graph, the fusion value for each time point is obtained by combining the differences between the two features and the standard value, which can integrate the abnormality of the two features.

[0022] In some embodiments of the present invention, in the step of calculating the weighted average of the difference between the normalized frequency value and the preset standard frequency value and the difference between the loudness value and the preset standard loudness value, the weighted average is calculated based on the following formula:

[0023]

[0024] Among them, Δ represents the weighted average value corresponding to a time point, A represents the normalized frequency value at the time point, and B represents the normalized loudness value at the time point. and They represent the weight values ​​corresponding to the normalized frequency value and the normalized loudness value respectively.

[0025] In some embodiments of the present invention, in the step of inputting the fusion analysis graph into a pre-trained neural network model and the neural network model outputting the fault probability, the fusion values ​​of multiple time points in the fusion analysis graph are constructed as an input vector, and the input vector is input into the pre-trained neural network model.

[0026] In the specific implementation process, the neural network model includes a convolutional layer, a fully connected layer and a classification layer. The classification layer outputs the abnormal value. This solution can jointly analyze the frequency and loudness over a period of time through the neural network model, and can more accurately determine whether an abnormal situation occurs.

[0027] Another aspect of the present invention relates to an electronic device having a computer program stored thereon, which is used to implement the steps of the aforementioned device status assessment method when the computer program is executed by a processor.

[0028] Another aspect of the present invention also relates to an equipment status assessment system, which includes a computer device, wherein the computer device includes a processor and a memory, wherein the memory stores computer instructions, and the processor is used to execute the computer instructions stored in the memory. When the computer instructions are executed by the processor, the system implements the steps implemented by the method.

[0029] In summary, the present invention has the following beneficial effects:

[0030] 1. This solution can first automatically count the power generation, and preliminarily determine the equipment status by comparing the predicted power generation with the actual power generation of the generator set in the predicted time period, and determine whether the equipment needs to be fault-checked. It can further analyze the sound frequency and loudness generated during the operation of the equipment through the audio data collected in the corresponding time period to determine whether the equipment is in a fault state;

[0031] 2. This solution can construct feature analysis graphs corresponding to the two features by statistically analyzing the frequency and loudness, and further combine the two features to construct a fusion analysis graph. In the fusion analysis graph, the fusion value at each time point is obtained by combining the difference between the two features and the standard value, which can integrate the abnormality of the two features;

[0032] 3. This solution can jointly analyze the frequency and loudness over a period of time through a neural network model, and can more accurately determine whether an abnormal situation occurs. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0034] Figure 1 A schematic diagram of an implementation of the device status assessment method of the present invention;

[0035] Figure 2 It is a schematic diagram of another implementation of the device status assessment method of the present invention. DETAILED DESCRIPTION

[0036] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Instead, they are merely examples of devices and methods consistent with some aspects of the present invention as detailed in the appended claims.

[0037] The terms used in the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The singular forms "a", "the" and "the" used in the present invention and the appended claims are also intended to include plural forms unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0038] like Figure 1 As shown, the present invention provides a method for evaluating equipment status, the steps of the method comprising:

[0039] Step S100, predicting the power generation in the prediction time period based on the historical power generation of the power generation set;

[0040] In the specific implementation process, the prediction model is used to predict the power generation in the prediction time period;

[0041] In some embodiments of the present invention, the prediction model is a long short-term memory network model (Long Short-Term Memory, LSTM) or a convolutional neural network model (Convolutional Neural Network, CNN);

[0042] LSTM controls the retention and forgetting of information between different moments by introducing three important gating mechanisms: Forget Gate, Input Gate, and Output Gate. These gating mechanisms allow the network to dynamically decide which information should be retained in memory and which should be forgotten, thus effectively solving the long-term dependency problem.

[0043] Forget Gate:

[0044] The forget gate determines what information to discard from the cell state.

[0045] It generates a value between 0 and 1 through a sigmoid function, indicating the degree of preservation of each state value.

[0046] Values ​​close to 0 indicate "forgetting", and values ​​close to 1 indicate "keeping".

[0047] Input Gate:

[0048] The input gate consists of two parts: a sigmoid layer and a tanh layer.

[0049] The sigmoid layer decides which values ​​will be updated into the cell state.

[0050] The tanh layer generates a new vector of candidate values.

[0051] These two outputs are multiplied to obtain the updated candidate value, which is then added to the cell state after processing by the forget gate to obtain the new cell state.

[0052] Output Gate:

[0053] The output gate determines the value of the next hidden state.

[0054] It first passes through a sigmoid layer to decide which cell states will be output.

[0055] Then, a tanh layer is used to generate candidate values ​​for the output state.

[0056] Finally, the output of the sigmoid layer is multiplied by the output of the tanh layer to form the final output.

[0057] Step S200, comparing the predicted power generation with the actual power generation of the generator set in the predicted time period to determine whether to perform a fault analysis on the generator set;

[0058] Step S300, if a fault analysis is performed on the generator set, a frequency analysis graph and a loudness analysis graph are respectively constructed based on the historical audio data of the generator set;

[0059] In a specific implementation process, the frequency analysis graph and the loudness analysis graph are both line statistical graphs.

[0060] Step S400, constructing a fusion analysis graph based on the frequency analysis graph and the loudness analysis graph, inputting the fusion analysis graph into a pre-trained neural network model, and the neural network model outputs a fault probability;

[0061] Step S500: determining whether the device state is a fault state based on the fault probability.

[0062] In a specific implementation process, in the step of determining whether the device state is a fault state based on the fault probability, if the fault probability is lower than a preset fault probability threshold, the device state is determined to be a fault state.

[0063] In some embodiments of the present invention, the fault probability threshold includes a first fault probability threshold and a second fault probability threshold, and the first fault probability threshold is greater than the second fault probability threshold. In the specific implementation process, if the fault probability is less than the first fault probability threshold and not less than the second fault probability threshold, the device state is the first fault state; if the fault probability is less than the second fault probability threshold, the device state is the second fault state. When the device state is the first fault state, a first instruction is issued to the control end, and when the device state is the second fault state, a second instruction is issued to the control end. The staff can determine the difference in device states by the different instructions obtained. When the device state is the second fault state, it means that the probability of device failure is high and needs to be processed as soon as possible. When the device state is the first fault state, you can choose to postpone the processing.

[0064] By adopting the above scheme, this scheme can first automatically count the power generation, and preliminarily determine the equipment status by comparing the predicted power generation with the actual power generation of the generator set in the predicted time period, and determine whether the equipment needs to be judged for faults. It can further analyze the audio data collected in the corresponding time period through the sound frequency and loudness generated during the operation of the equipment to determine whether the equipment is in a fault state.

[0065] In some embodiments of the present invention, in the step of determining whether the device state is a fault state based on the fault probability, if the device state is determined to be a fault state, historical audio data of the corresponding time period is recorded and uploaded.

[0066] By adopting the above scheme, the historical audio data of the corresponding time period can be used as a reference for the technicians, so that the technicians can analyze the fault through the historical audio data.

[0067] In some embodiments of the present invention, in the step of predicting the power generation in a prediction time period based on the power generation of the historical power generation of the generator set, the power generation at the same time point every day is constructed as a corresponding prediction vector, and the prediction vector is input into a preset power prediction model. The power prediction model outputs the power generation at the corresponding time point of the prediction day, and the predicted power generation at each time point in the prediction time period is determined in the same manner, and the average value of the power generation at each time point in the prediction time period is calculated as the power generation for the prediction time period.

[0068] In a specific implementation process, in the step of determining the predicted power generation at each time point in the prediction time period in the same manner, the predicted power generation at each time point in the prediction time period is determined by a power prediction model.

[0069] In some embodiments of the present invention, in the step of comparing the predicted power generation with the actual power generation of the generator set in the prediction time period to determine whether to perform a fault analysis on the generator set, if the predicted power generation calculated is less than the actual power generation of the generator set in the prediction time period, and the difference between the predicted power generation and the actual power generation of the generator set in the prediction time period is less than a preset power threshold, it is determined to perform a fault analysis on the generator set, otherwise, there is no need to perform a fault analysis on the generator set.

[0070] In the specific implementation process, in the step of comparing the predicted power generation with the actual power generation of the generator set in the predicted time period to determine whether to perform a fault analysis on the generator set, if the predicted power generation is at 6 pm, the actual power generation at 6 pm is compared with the predicted power generation at 6 pm.

[0071] In a specific implementation process, the power threshold may be a negative value.

[0072] Using the above scheme, the predicted power generation is compared with the actual power generation of the generator set during the predicted time period. If the actual power generation is too low, the generator set may have a fault and a fault analysis of the generator set is required.

[0073] In some embodiments of the present invention, in the steps of respectively constructing a frequency analysis graph and a loudness analysis graph based on the historical audio data of the generator set, a line graph with time as the horizontal axis and frequency value as the vertical axis is constructed based on the frequency data of the predicted time period; and a line graph with time as the horizontal axis and loudness value as the vertical axis is constructed based on the loudness data of the predicted time period.

[0074] In a specific implementation process, audio data corresponding to a time range of the predicted time period in the historical audio data is constructed.

[0075] In a specific implementation process, the historical audio data is collected by an audio device arranged at the generator set.

[0076] like Figure 2 As shown, in some embodiments of the present invention, the step of constructing a fusion analysis graph based on the frequency analysis graph and the loudness analysis graph includes:

[0077] Step S410, calculating the difference between the frequency value at each time point in the frequency analysis graph and the preset standard frequency value, and normalizing it;

[0078] Step S420, calculating the difference between the loudness value at each time point in the loudness analysis graph and the preset standard loudness value, and normalizing it;

[0079] In a specific implementation process, in the normalization step, the value is normalized to between -1 and 1.

[0080] Step S430, for each time point, calculate the weighted average of the difference between the normalized frequency value and the preset standard frequency value and the difference between the loudness value and the preset standard loudness value as the fusion value, and construct a line graph with the horizontal axis as time and the vertical axis as the calculated fusion value as the fusion analysis graph.

[0081] In a specific implementation process, the fusion analysis diagram combines the differences between the frequency characteristics and the loudness characteristics and the standard values, and represents the two differences in a fusion manner.

[0082] Step S440, inputting the fusion analysis graph into a pre-trained neural network model, and the neural network model outputs a fault probability.

[0083] In some embodiments of the present invention, the neural network model may be a convolutional neural network model or a graph neural network model. If it is a graph neural network model, the fusion analysis graph is directly input into the graph neural network model.

[0084] By adopting the above scheme, this scheme can construct feature analysis graphs corresponding to the two features respectively by statistically analyzing the frequency and loudness, and further combine the two features to construct a fusion analysis graph. In the fusion analysis graph, the fusion value for each time point is obtained by combining the differences between the two features and the standard value, which can integrate the abnormality of the two features.

[0085] In some embodiments of the present invention, in the step of calculating the weighted average of the difference between the normalized frequency value and the preset standard frequency value and the difference between the loudness value and the preset standard loudness value, the weighted average is calculated based on the following formula:

[0086]

[0087] Among them, Δ represents the weighted average value corresponding to a time point, A represents the normalized frequency value at the time point, and B represents the normalized loudness value at the time point. and They represent the weight values ​​corresponding to the normalized frequency value and the normalized loudness value respectively.

[0088] In some embodiments of the present invention, in the step of inputting the fusion analysis graph into a pre-trained neural network model and the neural network model outputting the fault probability, the fusion values ​​of multiple time points in the fusion analysis graph are constructed as an input vector, and the input vector is input into the pre-trained neural network model.

[0089] In a specific implementation process, if the neural network model is a convolutional neural network model, the fusion values ​​of multiple time points in the fusion analysis graph are constructed as an input vector, and the input vector is input into the pre-trained neural network model.

[0090] In the specific implementation process, if the neural network model is a convolutional neural network model, the neural network model includes a convolution layer, a fully connected layer and a classification layer. The abnormal value is output through the classification layer. This solution can jointly analyze the frequency and loudness over a period of time through the neural network model, and can more accurately determine whether an abnormal situation occurs.

[0091] Another aspect of the present invention relates to an electronic device having a computer program stored thereon, which, when executed by a processor, implements the steps of the aforementioned device status assessment method. The electronic device may be a tangible storage medium, such as a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a floppy disk, a hard disk, a removable storage disk, a CD-ROM, or any other form of storage medium known in the art.

[0092] An embodiment of the present invention also provides an equipment status assessment system, which includes a computer device, wherein the computer device includes a processor and a memory, wherein the memory stores computer instructions, and the processor is used to execute the computer instructions stored in the memory. When the computer instructions are executed by the processor, the system implements the steps implemented by the method.

[0093] It should be understood by those skilled in the art that the exemplary components, systems and methods described in conjunction with the embodiments disclosed herein can be implemented in hardware, software or a combination of the two. Whether it is performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present invention are programs or code segments used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link via a data signal carried in a carrier.

[0094] It should be clear that the present invention is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present invention.

[0095] In the present invention, features described and / or illustrated for one embodiment may be used in the same or similar manner in one or more other embodiments, and / or combined with features of other embodiments or replace features of other embodiments.

[0096] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the embodiments of the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for evaluating equipment status, characterized in that: The steps of the method include: Predicting the power generation during the forecast period based on the historical power generation of the power generating set; Compare the predicted power generation with the actual power generation of the generator set during the predicted time period to determine whether to perform a fault analysis on the generator set; If a fault analysis is performed on the generator set, a frequency analysis graph and a loudness analysis graph are respectively constructed based on the historical audio data of the generator set; Constructing a fusion analysis graph based on the frequency analysis graph and the loudness analysis graph, inputting the fusion analysis graph into a pre-trained neural network model, and the neural network model outputting a fault probability; It is determined whether the device state is a fault state based on the fault probability.

2. The device status assessment method according to claim 1, characterized in that: In the step of determining whether the device state is a fault state based on the fault probability, if the device state is determined to be a fault state, historical audio data of a corresponding time period is recorded and uploaded.

3. The device status assessment method according to claim 1, characterized in that: In the step of predicting the power generation in the prediction time period based on the power generation of the historical power generation of the power generator set, the power generation at the same time point every day is constructed as a corresponding prediction vector, and the prediction vector is input into a preset power prediction model. The power prediction model outputs the power generation at the corresponding time point of the prediction day, and the predicted power generation at each time point in the prediction time period is determined in the same way, and the average value of the power generation at each time point in the prediction time period is calculated as the power generation in the prediction time period.

4. The device status assessment method according to claim 3, characterized in that: In the step of comparing the predicted power generation with the actual power generation of the generator set in the prediction time period to determine whether to perform a fault analysis on the generator set, if the predicted power generation calculated is less than the actual power generation of the generator set in the prediction time period, and the difference between the predicted power generation and the actual power generation of the generator set in the prediction time period is less than a preset power threshold, it is determined to perform a fault analysis on the generator set, otherwise there is no need to perform a fault analysis on the generator set.

5. The device status assessment method according to any one of claims 1 to 4, characterized in that: In the step of constructing a frequency analysis graph and a loudness analysis graph respectively based on the historical audio data of the generator set, a line graph with time as the horizontal axis and frequency value as the vertical axis is constructed based on the frequency data of the predicted time period; and a line graph with time as the horizontal axis and loudness value as the vertical axis is constructed based on the loudness data of the predicted time period.

6. The device status assessment method according to claim 5, characterized in that: The step of constructing a fusion analysis graph based on the frequency analysis graph and the loudness analysis graph comprises: Calculate the difference between the frequency value at each time point in the frequency analysis graph and the preset standard frequency value, and normalize it; Calculate the difference between the loudness value at each time point in the loudness analysis graph and the preset standard loudness value, and normalize it; For each time point, the weighted average of the difference between the normalized frequency value and the preset standard frequency value and the difference between the loudness value and the preset standard loudness value is calculated, and a line graph with the horizontal axis as time and the vertical axis as the calculated weighted average is constructed as a fusion analysis graph.

7. The device status assessment method according to claim 6, characterized in that: In the step of calculating the weighted average of the difference between the normalized frequency value and the preset standard frequency value and the difference between the loudness value and the preset standard loudness value, the weighted average is calculated based on the following formula: Among them, Δ represents the weighted average value corresponding to a time point, A represents the normalized frequency value at the time point, and B represents the normalized loudness value at the time point. and They represent the weight values ​​corresponding to the normalized frequency value and the normalized loudness value respectively.

8. The device status assessment method according to claim 1, characterized in that: In the step of inputting the fusion analysis graph into a pre-trained neural network model and the neural network model outputting the fault probability, the fusion values ​​of multiple time points in the fusion analysis graph are constructed as an input vector, and the input vector is input into the pre-trained neural network model.

9. An electronic device, characterized in that: A computer program is stored thereon, and when the computer program is executed by a processor, the steps of the device status assessment method according to any one of claims 1 to 8 are implemented.

10. An equipment status assessment system, characterized in that: The system includes a computer device, which includes a processor and a memory, wherein the memory stores computer instructions, and the processor is used to execute the computer instructions stored in the memory. When the computer instructions are executed by the processor, the system implements the steps implemented by the device status assessment method described in any one of claims 1 to 8.

Images