Diesel generating set intelligent early warning system and method

Abstract

The invention provides an intelligent early warning system and method for a diesel generating set, and belongs to the technical field of diesel generating sets. The control panel is connected with the PLC; the communication module is connected with the PLC; the communication module is provided with a wired communication unit and a wireless communication unit; the analog-to-digital converter is connected with the PLC; the power supply module is connected with the PLC; the cloud server is connected with the communication module; the intelligent early warning module is connected with the communication module, the power supply module and the cloud server; the intelligent early warning module comprises an alarm unit which is connected with the analog-to-digital converter and the power supply module; the remote monitoring terminal is connected with the communication module; and the intelligent early warning terminal is connected with the cloud server. The method has the advantage that the reliability and convenience of operation and maintenance of the diesel generating set are greatly improved.

Description

Technical Field

[0001] The invention relates to the technical field of diesel generator sets, and in particular to an intelligent early warning system and method for diesel generator sets. Background Art

[0002] A diesel generator set is a small-scale power generation equipment, which refers to a power machine that uses diesel as fuel and a diesel engine as the prime mover to drive the generator to generate electricity; a complete diesel generator set is generally composed of a diesel engine, a generator, a control box, a fuel tank, a storage battery for starting and control, a protective device, an emergency cabinet and other components.

[0003] During the use of diesel generator sets, various mechanical or electrical faults are inevitable. Traditionally, faults are identified by installing indicator lights or instruments. However, if the diesel generator set itself is more complicated or frequently has multiple faults, this conventional alarm display method requires the installation of multiple indicator lights and instruments. At this time, experienced technicians are often required to identify and handle related faults, and manually record the handling experience. Due to the reliance on the experience of technicians, misjudgment is inevitable in the identification and handling of faults, and it is impossible to remotely monitor and manage faults, which brings great inconvenience to the operation and maintenance of diesel generator sets.

[0004] Therefore, how to provide an intelligent early warning system and method for diesel generator sets to improve the reliability and convenience of diesel generator set operation and maintenance has become a technical problem that needs to be solved urgently. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide an intelligent early warning system and method for a diesel generator set, so as to improve the reliability and convenience of the operation and maintenance of the diesel generator set.

[0006] In a first aspect, the present invention provides an intelligent early warning system for a diesel generator set, comprising:

[0007] 1PLC;

[0008] A control panel connected to the PLC;

[0009] a communication module connected to the PLC; the communication module is provided with a wired communication unit and a wireless communication unit;

[0010] an analog-to-digital converter connected to the PLC;

[0011] A power module connected to the PLC;

[0012] a cloud server connected to the communication module;

[0013] an intelligent early warning module, connected to the communication module, the power module and the cloud server;

[0014] The intelligent early warning module comprises:

[0015] an alarm unit connected to the analog-to-digital converter and the power module;

[0016] A remote monitoring terminal connected to the communication module;

[0017] An intelligent early warning terminal connected to the cloud server;

[0018] The alarm unit comprises:

[0019] an engine alarm subunit, connected to the analog-to-digital converter and the power module;

[0020] a generator alarm subunit connected to the analog-to-digital converter and the power module;

[0021] a fire alarm subunit, connected to the analog-to-digital converter and the power module;

[0022] A theft alarm subunit is connected to the analog-to-digital converter and the power module.

[0023] Furthermore, the control panel is a touch panel.

[0024] Furthermore, the wired communication unit is at least one of an RS485 communication unit, a CAN communication unit and an Ethernet communication unit.

[0025] Furthermore, the wireless communication unit is at least one of a 4G communication unit, a 5G communication unit, a WiFi communication unit, a Bluetooth communication unit, a Zigbee communication unit, an NFC communication unit, and a LoRa communication unit.

[0026] Furthermore, the engine alarm subunit includes:

[0027] At least one oil pressure sensor connected to the analog-to-digital converter and the power module;

[0028] At least one water temperature sensor connected to the analog-to-digital converter and the power module;

[0029] At least one intake and exhaust temperature sensor is connected to the analog-to-digital converter and the power module.

[0030] Furthermore, the generator alarm subunit includes:

[0031] At least one current transformer connected to the analog-to-digital converter and the power module;

[0032] At least one voltage transformer connected to the analog-to-digital converter and the power module;

[0033] At least one bearing temperature sensor connected to the analog-to-digital converter and the power module;

[0034] At least one differential sensor is connected to the analog-to-digital converter and the power module.

[0035] Furthermore, the fire alarm subunit includes:

[0036] At least one smoke sensor connected to the analog-to-digital converter and the power module;

[0037] At least one photoelectric sensor connected to the analog-to-digital converter and the power module;

[0038] At least one ambient temperature sensor is connected to the analog-to-digital converter and the power module.

[0039] Furthermore, the anti-theft alarm subunit includes:

[0040] At least one human proximity sensor connected to the analog-to-digital converter and the power module;

[0041] At least one laser sensor connected to the analog-to-digital converter and the power module;

[0042] At least one positioner connected to the analog-to-digital converter and the power module;

[0043] At least one vibration meter connected to the analog-to-digital converter and the power module;

[0044] At least one camera is connected to the communication module.

[0045] In a second aspect, the present invention provides an intelligent early warning method for a diesel generator set, comprising the following steps:

[0046] Step S1, the cloud server obtains a large amount of historical monitoring data collected by the alarm unit, performs data cleaning preprocessing on each of the historical monitoring data, and then annotates each of the preprocessed historical monitoring data with at least the fault type, fault level, and fault handling suggestion, and constructs a data set based on each of the annotated historical monitoring data;

[0047] Step S2: creating a fault diagnosis model based on a deep neural network, training the fault diagnosis model using the data set, and deploying the trained fault diagnosis model to an intelligent early warning terminal;

[0048] Step S3, after the early warning system is powered on by the power module, the real-time monitoring data collected by the alarm unit is converted into digital form by the analog-to-digital converter and then sent to the PLC in real time;

[0049] Step S4, the PLC performs a preliminary diagnosis on the real-time monitoring data based on the pre-stored historical monitoring data to obtain a preliminary diagnosis result, displays the preliminary diagnosis result through a control panel, and sends the real-time monitoring data to a remote monitoring terminal and a cloud server through a communication module;

[0050] Step S5: The cloud server stores the received real-time monitoring data and sends the real-time monitoring data to the intelligent early warning terminal;

[0051] Step S6: The intelligent early warning terminal inputs the received real-time monitoring data into a fault diagnosis model to obtain a fault diagnosis report, and pushes the fault diagnosis report to a pre-associated management terminal in real time for early warning;

[0052] Step S7: The management terminal accesses the cloud server to query the real-time monitoring data.

[0053] Furthermore, the step S2 is specifically as follows:

[0054] A fault diagnosis model is created based on a deep neural network. The data set is divided into a training set, a validation set and a test set based on a preset ratio. The fault diagnosis model is trained with the training set until the loss function of the fault diagnosis model is less than a preset loss threshold. The trained fault diagnosis model is then verified with the validation set to determine whether the diagnostic accuracy is greater than the preset accuracy threshold. If not, the verification fails and the data set is expanded to continue training. If yes, the verification passes, and:

[0055] The test set is used to test the verified fault diagnosis model to determine whether the confidence of the fault diagnosis model is greater than a preset confidence threshold. If not, the test fails and the data set is expanded to continue training; if so, the test passes, the training is terminated, and the fault diagnosis model that has passed the test is deployed to the intelligent early warning terminal.

[0056] The advantages of the present invention are:

[0057] By setting PLC, control panel, communication module, analog-to-digital converter, power module, cloud server, and intelligent early warning module, the intelligent early warning module includes an alarm unit, a remote monitoring terminal, and an intelligent early warning terminal. The PLC is connected to the alarm unit through the analog-to-digital converter, and is also connected to the control panel and the communication module. The power module is connected to the PLC and the intelligent alarm module respectively. The communication module is connected to the cloud server, the remote monitoring terminal, and the camera of the alarm unit respectively, and the cloud server is connected to the intelligent early warning terminal. Since the alarm unit is provided with an engine alarm subunit, a generator alarm subunit, a fire alarm subunit, and an anti-theft alarm subunit, each subunit is provided with different types of sensors, which can collect multi-dimensional monitoring data. The PLC can generate a preliminary diagnosis result of the fault by comparing the real-time monitoring data and the historical monitoring data and display it on the control panel. The real-time monitoring data can also be uploaded to the cloud server for further analysis to ensure the accuracy of the fault diagnosis. The remote monitoring terminal can receive the real-time monitoring data in real time through the communication module. The intelligent early warning terminal can issue an early warning based on the real-time monitoring data sent by the cloud server, that is, remote monitoring and early warning are realized, which ultimately greatly improves the reliability and convenience of the operation and maintenance of the diesel generator set. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The present invention will be further described below in conjunction with embodiments with reference to the accompanying drawings.

[0059] Figure 1 The invention discloses a circuit principle block diagram of an intelligent early warning system for a diesel generator set.

[0060] Figure 2 The present invention is a flowchart of an intelligent early warning method for a diesel generator set.

[0061] Marking Description:

[0062] 100-an intelligent early warning system for a diesel generator set, 1-PLC, 2-control panel, 3-communication module, 4-analog-to-digital converter, 5-power module, 6-cloud server, 7-intelligent early warning module, 31-wired communication unit, 32-wireless communication unit, 71-alarm unit, 72-remote monitoring terminal, 73-intelligent early warning terminal, 711-engine alarm subunit, 712-generator alarm subunit, 713-fire alarm subunit, 714-theft alarm subunit, 7111-Oil pressure sensor, 7112-Water temperature sensor, 7113-Intake and exhaust temperature sensor, 7121-Current transformer, 7122-Voltage transformer, 7123-Bearing temperature sensor, 7124-Differential sensor, 7131-Smoke sensor, 7132-Photoelectric sensor, 7133-Ambient temperature sensor, 7141-Human proximity sensor, 7142-Laser sensor, 7143-Positioner, 7144-Vibrator, 7145-Camera. DETAILED DESCRIPTION

[0063] The technical solution in the embodiment of the present application has the following overall idea: an alarm unit 71 including an engine alarm subunit 711, a generator alarm subunit 712, a fire alarm subunit 713 and an anti-theft alarm subunit 714 is set, and each subunit is provided with different types of sensors, which can collect multi-dimensional monitoring data. PLC1 can generate a preliminary diagnosis result of the fault by comparing the real-time monitoring data and historical monitoring data collected by the alarm unit and display it on the control panel 2. The real-time monitoring data can also be uploaded to the cloud server 6 for further analysis to ensure the accuracy of the fault diagnosis. The remote monitoring terminal 72 can receive the real-time monitoring data in real time through the communication module 3, and the intelligent early warning terminal 73 can issue an early warning based on the real-time monitoring data sent by the cloud server 6, that is, remote monitoring and early warning are realized to improve the reliability and convenience of the operation and maintenance of the diesel generator set.

[0064] Please refer to Figure 1 to Figure 2 As shown, a preferred embodiment of an intelligent early warning system 100 for a diesel generator set of the present invention comprises:

[0065] A PLC 1, used to control the operation of the early warning system 100;

[0066] a control panel 2, connected to the PLC1, for sending control instructions to the PLC1 and displaying relevant monitoring data or preliminary diagnosis results of faults;

[0067] A communication module 3 connected to the PLC1; the communication module 3 is provided with a wired communication unit 31 and a wireless communication unit 32, which are used for the communication between the PLC1 and the cloud server 6, the remote monitoring terminal 72 and the camera 7145;

[0068] an analog-to-digital converter 4, connected to the PLC 1, for converting the analog signal collected by the alarm unit 71 into a digital signal;

[0069] a power module 5, connected to the PLC 1, for supplying power to the early warning system 100;

[0070] a cloud server 6, connected to the communication module 3, for storing, forwarding and querying the real-time monitoring data of the alarm unit 71;

[0071] an intelligent early warning module 7, connected to the communication module 3, the power module 5 and the cloud server 6, for collecting the operation data and environmental data of the diesel generator set (not shown), and issuing an early warning when a fault occurs in the diesel generator set;

[0072] The intelligent early warning module 7 includes:

[0073] an alarm unit 71, connected to the analog-to-digital converter 4 and the power module 5, and disposed inside or around the diesel generator set, for safety monitoring of the diesel generator set;

[0074] a remote monitoring terminal 72 connected to the communication module 4 and used to remotely view the real-time monitoring data collected by the alarm unit 71;

[0075] An intelligent early warning terminal 73, connected to the cloud server 6, for early warning when a diesel generator set fails;

[0076] The alarm unit 71 comprises:

[0077] An engine alarm subunit 711, connected to the analog-to-digital converter 4 and the power module 5, and arranged on the engine of the diesel generator set;

[0078] a generator alarm subunit 712, connected to the analog-to-digital converter 4 and the power module 5, and arranged on the generator of the diesel generator set;

[0079] a fire alarm subunit 713, connected to the analog-to-digital converter 4 and the power module 5;

[0080] An anti-theft alarm subunit 714 is connected to the analog-to-digital converter 4 and the power module 5 .

[0081] The control panel 2 is a touch panel.

[0082] The wired communication unit 31 is at least one of an RS485 communication unit, a CAN communication unit and an Ethernet communication unit.

[0083] The wireless communication unit 32 is at least one of a 4G communication unit, a 5G communication unit, a WiFi communication unit, a Bluetooth communication unit, a Zigbee communication unit, an NFC communication unit, and a LoRa communication unit.

[0084] The engine alarm subunit 711 includes:

[0085] At least one oil pressure sensor 7111, connected to the analog-to-digital converter 4 and the power module 5;

[0086] At least one water temperature sensor 7112, connected to the analog-to-digital converter 4 and the power module 5;

[0087] At least one intake and exhaust temperature sensor 7113 is connected to the analog-to-digital converter 4 and the power module 5.

[0088] The generator alarm subunit 712 includes:

[0089] At least one current transformer 7121, connected to the analog-to-digital converter 4 and the power module 5;

[0090] At least one voltage transformer 7122, connected to the analog-to-digital converter 4 and the power module 5;

[0091] At least one bearing temperature sensor 7123, connected to the analog-to-digital converter 4 and the power module 5;

[0092] At least one differential sensor 7124 is connected to the analog-to-digital converter 4 and the power module 5 .

[0093] The fire alarm subunit 713 includes:

[0094] At least one smoke sensor 7131, connected to the analog-to-digital converter 4 and the power module 5;

[0095] At least one photoelectric sensor 7132, connected to the analog-to-digital converter 4 and the power module 5;

[0096] At least one ambient temperature sensor 7133 is connected to the analog-to-digital converter 4 and the power module 5 .

[0097] The anti-theft alarm subunit 714 includes:

[0098] At least one human proximity sensor 7141, connected to the analog-to-digital converter 4 and the power module 5;

[0099] At least one laser sensor 7142, connected to the analog-to-digital converter 4 and the power module 5;

[0100] At least one positioner 7143, connected to the analog-to-digital converter 4 and the power module 5;

[0101] At least one vibration meter 7144 connected to the analog-to-digital converter 4 and the power module 5;

[0102] At least one camera 7145 is connected to the communication module 3 .

[0103] A preferred embodiment of an intelligent early warning method for a diesel generator set of the present invention comprises the following steps:

[0104] Step S1, the cloud server obtains a large amount of historical monitoring data collected by the alarm unit, performs data cleaning preprocessing on each of the historical monitoring data, and then annotates each of the preprocessed historical monitoring data with at least the fault type, fault level, and fault handling suggestion, and constructs a data set based on each of the annotated historical monitoring data;

[0105] Step S2: creating a fault diagnosis model based on a deep neural network, training the fault diagnosis model using the data set, and deploying the trained fault diagnosis model to an intelligent early warning terminal;

[0106] Step S3, after the early warning system is powered on by the power module, the real-time monitoring data collected by the alarm unit is converted into digital form by the analog-to-digital converter and then sent to the PLC in real time;

[0107] Step S4, the PLC performs a preliminary diagnosis on the real-time monitoring data based on the pre-stored historical monitoring data to obtain a preliminary diagnosis result, displays the preliminary diagnosis result through the control panel, and sends the real-time monitoring data to the remote monitoring terminal and the cloud server through the communication module; that is, the PLC compares the historical monitoring data with the real-time monitoring data to determine whether the value of the real-time monitoring data is abnormal to perform a preliminary diagnosis; in specific implementation, the preliminary diagnosis result can be classified into three levels: general indication, general alarm, and fault shutdown;

[0108] Step S5: The cloud server stores the received real-time monitoring data and sends the real-time monitoring data to the intelligent early warning terminal;

[0109] Step S6: the intelligent early warning terminal inputs the received real-time monitoring data into a fault diagnosis model to obtain a fault diagnosis report, and pushes the fault diagnosis report to a pre-associated management terminal in real time for early warning; the intelligent early warning terminal continuously optimizes the fault diagnosis model based on the fault diagnosis report;

[0110] Step S7: The management terminal accesses the cloud server to query the real-time monitoring data.

[0111] Furthermore, the step S2 is specifically as follows:

[0112] A fault diagnosis model is created based on a deep neural network. The data set is divided into a training set, a validation set and a test set based on a preset ratio. The fault diagnosis model is trained with the training set until the loss function of the fault diagnosis model is less than a preset loss threshold. The trained fault diagnosis model is then verified with the validation set to determine whether the diagnostic accuracy is greater than the preset accuracy threshold. If not, the verification fails and the data set is expanded to continue training. If yes, the verification passes, and:

[0113] The fault diagnosis model that has passed the verification is tested using the test set to determine whether the confidence of the fault diagnosis model is greater than the preset confidence threshold. If not, the test fails and the data set is expanded to continue training; if yes, the test passes, the training ends, and the fault diagnosis model that has passed the test is deployed to the intelligent early warning terminal. By verifying and testing the trained fault diagnosis model, the diagnostic accuracy of the fault diagnosis model is greatly improved.

[0114] In step S5, the real-time monitoring data received and stored by the cloud server is specifically:

[0115] The cloud server converts the real-time monitoring data from plain text to hexadecimal data, swaps the number 1 and the letter I in the hexadecimal data, and swaps the number 2 and the letter Z to obtain the first-level encrypted data, encrypts the first-level encrypted data into the second-level encrypted data through the 3DES algorithm, shifts each character of the second-level encrypted data 6 bits to the left in a loop to obtain the third-level encrypted data, encrypts the third-level encrypted data into encrypted data through the RC6 algorithm, stores the encrypted data to a designated path, and performs distributed backup of the encrypted data. By performing multi-level encryption and distributed backup of the real-time monitoring data, the security of the real-time monitoring data storage is effectively improved, and the real-time monitoring data is prevented from being stolen and tampered with in plain text.

[0116] In summary, the advantages of the present invention are:

[0117] By setting PLC, control panel, communication module, analog-to-digital converter, power module, cloud server, and intelligent early warning module, the intelligent early warning module includes an alarm unit, a remote monitoring terminal, and an intelligent early warning terminal. The PLC is connected to the alarm unit through the analog-to-digital converter, and is also connected to the control panel and the communication module. The power module is connected to the PLC and the intelligent alarm module respectively. The communication module is connected to the cloud server, the remote monitoring terminal, and the camera of the alarm unit respectively, and the cloud server is connected to the intelligent early warning terminal. Since the alarm unit is provided with an engine alarm subunit, a generator alarm subunit, a fire alarm subunit, and an anti-theft alarm subunit, each subunit is provided with different types of sensors, which can collect multi-dimensional monitoring data. The PLC can generate a preliminary diagnosis result of the fault by comparing the real-time monitoring data and the historical monitoring data and display it on the control panel. The real-time monitoring data can also be uploaded to the cloud server for further analysis to ensure the accuracy of the fault diagnosis. The remote monitoring terminal can receive the real-time monitoring data in real time through the communication module. The intelligent early warning terminal can issue an early warning based on the real-time monitoring data sent by the cloud server, that is, remote monitoring and early warning are realized, which ultimately greatly improves the reliability and convenience of the operation and maintenance of the diesel generator set.

[0118] Although the specific implementation modes of the present invention are described above, those skilled in the art should understand that the specific implementation modes described are only illustrative and are not intended to limit the scope of the present invention. Equivalent modifications and changes made by those skilled in the art in accordance with the spirit of the present invention should be included in the scope of protection of the claims of the present invention.

Claims

1. An intelligent early warning system for diesel generator sets, characterized in that: include: 1PLC; A control panel connected to the PLC; a communication module connected to the PLC; the communication module is provided with a wired communication unit and a wireless communication unit; an analog-to-digital converter connected to the PLC; A power module connected to the PLC; a cloud server connected to the communication module; an intelligent early warning module, connected to the communication module, the power module and the cloud server; The intelligent early warning module comprises: an alarm unit connected to the analog-to-digital converter and the power module; A remote monitoring terminal connected to the communication module; An intelligent early warning terminal connected to the cloud server; The alarm unit comprises: an engine alarm subunit, connected to the analog-to-digital converter and the power module; a generator alarm subunit, connected to the analog-to-digital converter and the power module; a fire alarm subunit, connected to the analog-to-digital converter and the power module; A theft alarm subunit is connected to the analog-to-digital converter and the power module.

2. The intelligent early warning system for diesel generator sets according to claim 1, characterized in that: The control panel is a touch panel.

3. The intelligent early warning system for diesel generator sets according to claim 1, characterized in that: The wired communication unit is at least one of an RS485 communication unit, a CAN communication unit and an Ethernet communication unit.

4. The intelligent early warning system for diesel generator sets according to claim 1, characterized in that: The wireless communication unit is at least one of a 4G communication unit, a 5G communication unit, a WiFi communication unit, a Bluetooth communication unit, a Zigbee communication unit, an NFC communication unit, and a LoRa communication unit.

5. The intelligent early warning system for diesel generator sets according to claim 1, characterized in that: The engine alarm subunit comprises: At least one oil pressure sensor connected to the analog-to-digital converter and the power module; At least one water temperature sensor connected to the analog-to-digital converter and the power module; At least one intake and exhaust temperature sensor is connected to the analog-to-digital converter and the power module.

6. The intelligent early warning system for diesel generator sets according to claim 1, characterized in that: The generator alarm subunit comprises: At least one current transformer connected to the analog-to-digital converter and the power module; At least one voltage transformer connected to the analog-to-digital converter and the power module; At least one bearing temperature sensor connected to the analog-to-digital converter and the power module; At least one differential sensor is connected to the analog-to-digital converter and the power module.

7. The intelligent early warning system for diesel generator sets according to claim 1, characterized in that: The fire alarm subunit comprises: At least one smoke sensor connected to the analog-to-digital converter and the power module; At least one photoelectric sensor connected to the analog-to-digital converter and the power module; At least one ambient temperature sensor is connected to the analog-to-digital converter and the power module.

8. The intelligent early warning system for diesel generator sets according to claim 1, characterized in that: The anti-theft alarm subunit comprises: At least one human proximity sensor connected to the analog-to-digital converter and the power module; At least one laser sensor connected to the analog-to-digital converter and the power module; At least one positioner connected to the analog-to-digital converter and the power module; At least one vibration meter connected to the analog-to-digital converter and the power module; At least one camera is connected to the communication module.

9. An intelligent early warning method for a diesel generator set, characterized in that: The method requires the use of the intelligent early warning system for diesel generator sets as claimed in any one of claims 1 to 8, comprising the following steps: Step S1, the cloud server obtains a large amount of historical monitoring data collected by the alarm unit, performs data cleaning preprocessing on each of the historical monitoring data, and then annotates each of the preprocessed historical monitoring data with at least the fault type, fault level, and fault handling suggestion, and constructs a data set based on each of the annotated historical monitoring data; Step S2: creating a fault diagnosis model based on a deep neural network, training the fault diagnosis model using the data set, and deploying the trained fault diagnosis model to an intelligent early warning terminal; Step S3, after the early warning system is powered on by the power module, the real-time monitoring data collected by the alarm unit is converted into digital form by the analog-to-digital converter and then sent to the PLC in real time; Step S4, the PLC performs a preliminary diagnosis on the real-time monitoring data based on the pre-stored historical monitoring data to obtain a preliminary diagnosis result, displays the preliminary diagnosis result through a control panel, and sends the real-time monitoring data to a remote monitoring terminal and a cloud server through a communication module; Step S5: The cloud server stores the received real-time monitoring data and sends the real-time monitoring data to the intelligent early warning terminal; Step S6: The intelligent early warning terminal inputs the received real-time monitoring data into a fault diagnosis model to obtain a fault diagnosis report, and pushes the fault diagnosis report to a pre-associated management terminal in real time for early warning; Step S7: The management terminal accesses the cloud server to query the real-time monitoring data.

10. The intelligent early warning method for a diesel generator set according to claim 9, characterized in that: The step S2 is specifically as follows: A fault diagnosis model is created based on a deep neural network. The data set is divided into a training set, a validation set and a test set based on a preset ratio. The fault diagnosis model is trained with the training set until the loss function of the fault diagnosis model is less than a preset loss threshold. The trained fault diagnosis model is then verified with the validation set to determine whether the diagnostic accuracy is greater than the preset accuracy threshold. If not, the verification fails and the data set is expanded to continue training. If yes, the verification passes, and: The test set is used to test the verified fault diagnosis model to determine whether the confidence of the fault diagnosis model is greater than a preset confidence threshold. If not, the test fails and the data set is expanded to continue training; if so, the test passes, the training is terminated, and the fault diagnosis model that has passed the test is deployed to the intelligent early warning terminal.

Images