Risk assessment fault diagnosis and fault self-healing method based on cems system

Abstract

The invention belongs to the technical field of cems systems, and discloses a risk assessment fault diagnosis and fault self-healing method based on a cems system, the method realizes fault diagnosis and self-healing based on a machine learning mechanism, and the method comprises the following specific steps: step 1, environment monitoring and data uploading; step 2, remote control and feedback; 3, training a machine learning model; step 4, establishing a pre-alarm module; and 5, establishing a self-healing module. Through real-time monitoring, data acquisition and fault diagnosis of a machine learning model, the reliability and stability of the CEMS system are greatly improved, possible problems of equipment are accurately predicted through deep learning and machine learning technologies, the faults are accurately diagnosed, and compared with a traditional method, the method has the advantages that the efficiency is high, and the cost is low. The intelligent diagnosis mechanism can effectively reduce the manual intervention demand, greatly improve the working efficiency, and reduce the equipment downtime and the manual maintenance cost.

Description

Technical Field

[0001] The invention belongs to the technical field of cems systems, and in particular relates to a method for risk assessment, fault diagnosis and fault self-healing based on a cems system. Background Art

[0002] The CEMS dynamic environment monitoring system, that is, the power and environment monitoring system in the continuous emission and environmental monitoring system, is mainly used for real-time monitoring of various equipment and environmental parameters in the computer room or other facilities, and usually includes multiple monitoring devices and sensors. Through these devices, the system collects, processes and transmits data, and then sends the monitoring results to the host server.

[0003] However, in actual applications, the CEMS system may encounter various faults, such as sensor failure, data transmission interruption, software abnormality, etc., and maintenance is required at this time. When an existing CEMS system fails, the sensor data triggers the threshold, the host issues an alarm, and the maintenance personnel are notified through text messages, phone calls, or voice broadcasts. However, only an alarm at a certain measurement point is not informed, and the real problem behind the alarm at that measurement point cannot be analyzed. It is impossible to summarize and analyze the cause of the alarm from the overall perspective of the computer room, and analyze whether the current alarm measurement point is affected by the operation of other equipment to cause the alarm, and whether it will affect other equipment. It is also impossible to provide practical repairs and solutions, which will cause negligence in maintenance work and result in the failure to discover and handle faults in a timely manner.

[0004] Moreover, before triggering an alarm, the existing CEMS system cannot give a risk warning when the data of the device sensor is close to the safety threshold. When the data is unstable and fluctuates greatly, it may cause frequent alarms. Therefore, a method for fault diagnosis and self-healing of the CEMS system is needed to optimize it. Summary of the invention

[0005] The object of the present invention is to provide a method for risk assessment fault diagnosis and fault self-healing based on a CEMS system to solve the problems raised in the above background technology.

[0006] In order to achieve the above object, the present invention provides the following technical solution: a method for risk assessment fault diagnosis and fault self-healing based on a CEMS system, wherein the method realizes fault diagnosis and self-healing based on a machine learning mechanism, and the specific steps are as follows:

[0007] Step 1: Environmental monitoring and data upload: deploy sensors for monitoring environmental parameters and upload environmental data to the host;

[0008] Step 2: Remote control and feedback: Build a remote control and feedback system to allow technicians to remotely control the facilities and equipment in the computer room through a central monitoring platform;

[0009] Step 3: Machine learning model training: Process the large amount of historical operation data collected in the CEMS system and train the model with the processed data so that it can automatically identify potential failure modes.

[0010] Step 4: Establish a warning module. Construct a warning module so that it can detect potential failure risks in advance and issue warning signals through real-time monitoring and analysis of the CEMS system;

[0011] Step 5: The self-healing module is established. When the model detects a fault in the system, the self-healing module will start the corresponding self-healing program to repair and adjust the system.

[0012] Preferably, the machine learning model can analyze system operation data in real time, automatically identify potential failure modes, and improve the accuracy of fault diagnosis through multi-parameter fusion analysis.

[0013] Preferably, the training method of the machine learning model is:

[0014] A1, collect a large amount of historical operation data from the CEMS system and pre-process the data;

[0015] A2, extract useful feature information from the preprocessed data, and this feature information will be used as the input vector of the machine learning model;

[0016] A3, selecting a suitable machine learning model, and the machine learning model is one of support vector machine (SVM), random forest (RF) and neural network (NN);

[0017] A4, use the training data set to train the selected model and optimize the model parameters through methods such as cross-validation;

[0018] A5, inputs the real-time collected data into the trained machine learning model and outputs the fault diagnosis results.

[0019] Preferably, during the environmental monitoring and data uploading stage, the sensors deployed in the computer room upload the data to the host quickly and securely through the protocol, and the protocol needs to support breakpoint resumption and data encryption functions to ensure the integrity and security of the data.

[0020] Preferably, the remote control and feedback system allows technicians to remotely control the facilities and equipment in the computer room through the central monitoring platform, and provide real-time feedback on the operating status and operation results of the equipment.

[0021] Preferably, the steps of realizing the early warning function of the pre-warning module are:

[0022] B1, first analyze the historical data, identify the long-term change trend and abnormal points of the data, and combine the trend analysis results with the current data to evaluate the potential failure risk level;

[0023] B2, when the assessment is completed and the assessment results show that there is a high risk of failure, an early warning signal is generated and notified to relevant personnel through interface display and email notification;

[0024] B3, the early warning module will provide corresponding preventive measures based on different types of potential failure risks, helping users take timely measures to reduce the possibility of failure.

[0025] Preferably, the self-healing module includes a self-healing strategy library, a self-healing execution unit and a self-healing effect evaluation unit;

[0026] The specific levels of the self-healing execution library are:

[0027] C1, the basic layer, contains basic self-healing operations at the device level;

[0028] C2, the environmental layer, includes self-healing strategies for temperature and humidity, air conditioning, and fresh air environment control equipment;

[0029] C3, the network layer, includes self-healing solutions related to network communication and data transmission;

[0030] C4, application layer: self-healing measures for software systems and data processing, and each policy item includes: fault type identification, fault severity rating, self-healing operation step sequence, operation timeout limit, expected effect indicator and historical execution success rate.

[0031] Preferably, the self-healing execution unit includes a strategy matching layer, an execution control layer and an emergency processing layer;

[0032] The strategy matching layer will take multi-dimensional factors into consideration and retrieve the self-healing solution with the highest matching degree from the strategy library based on the fault diagnosis results; the execution control layer will establish an operation queue, execute the self-healing steps in order of priority, and set up an operation timeout mechanism to prevent deadlock; the emergency processing layer will set up a rollback mechanism to restore the system to its original state when self-healing fails, and automatically upgrade the alarm level when the self-healing failure reaches a preset number of times.

[0033] Preferably, the self-healing effect evaluation unit includes a real-time evaluation function, an effect analysis function and a strategy optimization function.

[0034] Preferably, the real-time evaluation function will monitor changes in key performance indicators and compare the system status before and after the failure to verify whether the expected repair effect is achieved; the effect analysis function will calculate the success rate of the self-healing operation and calculate the average repair time. When the repair fails, the cause of the failure will be analyzed and a report will be generated; the policy optimization function will update the policy library according to the evaluation results and adjust the policy priority and self-healing parameters.

[0035] The beneficial effects of the present invention are as follows:

[0036] 1. The present invention greatly improves the reliability and stability of the CEMS system through real-time monitoring, data collection and fault diagnosis of machine learning models, so that the system can automatically detect abnormal changes in equipment operation and identify potential faults in a timely manner. Through deep learning and machine learning technology, the system can analyze based on historical data and real-time data, accurately predict possible problems of the equipment, and accurately diagnose faults. Compared with traditional methods, this intelligent diagnosis mechanism can effectively reduce the need for manual intervention, avoid misdiagnosis or delayed response caused by human factors, greatly improve work efficiency, reduce equipment downtime and manual maintenance costs, thereby ensuring efficient operation of the system.

[0037] 2. The present invention can automatically execute preset repair strategies through the setting of self-healing modules, effectively reducing the system downtime and human intervention, and the self-healing strategy library contains most of the repair solutions. The system can select the most appropriate repair measures from the strategy library according to the type, severity and historical repair data of the fault, and automatically start and execute them, so that the system can quickly resume normal operation without relying on manual labor. More importantly, after the self-healing operation is completed, the system will automatically evaluate the repair effect to ensure that with the accumulation of experience, the system's repair efficiency and success rate will continue to improve, reducing equipment maintenance costs and improving operation and maintenance efficiency.

[0038] 3. The present invention integrates and analyzes equipment operation data, environmental data, and historical fault data through multi-source data fusion, so that the system can identify potential risks in advance before a major equipment failure occurs, and predict the changing trends of key parameters through a time series analysis algorithm, thereby realizing an early warning function. Further, through this active prevention and control strategy, the staff can take effective measures before the failure occurs, greatly reducing the frequency of equipment failures and improving the reliability and safety of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 This is a functional framework diagram of the machine learning model of the present invention;

[0040] Figure 2 This is a functional schematic diagram of the self-healing module of the present invention. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0042] like Figure 1 to Figure 2 As shown, the embodiment of the present invention provides a method for risk assessment fault diagnosis and fault self-healing based on a CEMS system. The method implements fault diagnosis and self-healing based on a machine learning mechanism, and the specific steps are as follows:

[0043] Step 1: Environmental monitoring and data upload: deploy sensors for monitoring environmental parameters and upload environmental data to the host;

[0044] Real-time monitoring of environmental parameters in the equipment room, such as temperature, humidity, smoke, and water immersion, by deploying a variety of sensors and configuring backup monitoring equipment;

[0045] Step 2: Remote control and feedback: Build a remote control and feedback system to allow technicians to remotely control the facilities and equipment in the computer room through a central monitoring platform;

[0046] Step 3: Machine learning model training: Process the large amount of historical operation data collected in the CEMS system and train the model with the processed data so that it can automatically identify potential failure modes.

[0047] Step 4: Establish a warning module. The warning module can detect potential failure risks in advance through real-time monitoring and analysis of the CEMS system, and issue early warning signals so that preventive measures can be taken in time to avoid failures.

[0048] Step 5: The self-healing module is established. When the model detects a fault in the system, the self-healing module will start the corresponding self-healing program to repair and adjust the system.

[0049] Through real-time monitoring, data collection and fault diagnosis based on machine learning models, the reliability and stability of the CEMS system have been greatly improved, enabling the system to automatically detect abnormal changes in equipment operation and identify potential faults in a timely manner. Through deep learning and machine learning technology, the system can analyze historical data and real-time data, accurately predict possible problems with the equipment, and accurately diagnose faults. Compared with traditional methods, this intelligent diagnostic mechanism can effectively reduce the need for manual intervention, avoid misdiagnosis or delayed response caused by human factors, greatly improve work efficiency, reduce equipment downtime and manual maintenance costs, thereby ensuring efficient operation of the system.

[0050] Through the setting of the self-healing module, it can automatically execute the preset repair strategy, effectively reducing the system downtime and human intervention. The self-healing strategy library contains most of the repair solutions. The system can select the most appropriate repair measures from the strategy library according to the type, severity and historical repair data of the fault, and automatically start and execute them, so that the system can quickly resume normal operation without relying on manual labor. More importantly, after the self-healing operation is completed, the system will automatically evaluate the repair effect to ensure that with the accumulation of experience, the system's repair efficiency and success rate will continue to improve, reducing equipment maintenance costs and improving operation and maintenance efficiency.

[0051] Among them, the machine learning model can analyze system operation data in real time, automatically identify potential fault modes, and improve the accuracy of fault diagnosis through multi-parameter fusion analysis (such as temperature, humidity, wind speed, etc.).

[0052] The system can automatically complete fault detection, diagnosis and partial repair work, greatly reducing the need for human intervention.

[0053] Among them, the training method of the machine learning model is:

[0054] A1, collect a large amount of historical operation data from the CEMS system and preprocess the data (including filling missing values, normalization, etc.);

[0055] Fusion prediction of multi-source data integrates multi-dimensional information such as environmental data, equipment operation data, historical fault data, and establishes a deep learning model to achieve equipment life prediction and performance degradation analysis. Then, through the time series analysis algorithm, the changing trends of key parameters are predicted.

[0056] A2, extract useful feature information from the preprocessed data, and this feature information will be used as the input vector of the machine learning model;

[0057] These characteristic information include temperature change rate and electrical energy voltage range;

[0058] A3, selecting a suitable machine learning model, and the machine learning model is one of support vector machine (SVM), random forest (RF) and neural network (NN);

[0059] A4, use the training data set to train the selected model and optimize the model parameters through methods such as cross-validation;

[0060] A5, inputs the real-time collected data into the trained machine learning model and outputs the fault diagnosis results.

[0061] The prediction results of the model make it easier to predict and warn about the future operating status of the system.

[0062] Among them, during the environmental monitoring and data uploading stage, the sensors deployed in the computer room upload the data to the host quickly and securely through the protocol, and the protocol must support breakpoint resumption and data encryption functions to ensure the integrity and security of the data.

[0063] Among them, the remote control and feedback system allows technicians to remotely control the facilities and equipment in the computer room (such as air conditioners, dehumidifiers, lighting, etc.) through the central monitoring platform, and provide real-time feedback on the operating status and operation results of the equipment.

[0064] The steps of realizing the early warning function of the pre-warning module are as follows:

[0065] B1, first analyze the historical data, identify the long-term change trend and abnormal points of the data, and combine the trend analysis results with the current data to evaluate the potential failure risk level;

[0066] B2, when the assessment is completed and the assessment results show that there is a high risk of failure, an early warning signal is generated and notified to relevant personnel through interface display and email notification;

[0067] B3, the early warning module will provide corresponding preventive measures based on different types of potential failure risks, helping users take timely measures to reduce the possibility of failure.

[0068] By integrating and analyzing multi-source data, equipment operation data, environmental data, and historical fault data, the system can identify potential risks in advance before a major equipment failure occurs, and predict the changing trends of key parameters through a time series analysis algorithm, thereby realizing an early warning function. Further, through this proactive prevention and control strategy, staff can take effective measures before a failure occurs, greatly reducing the frequency of equipment failures and improving the reliability and safety of the system.

[0069] The self-healing module includes a self-healing strategy library, a self-healing execution unit and a self-healing effect evaluation unit;

[0070] The specific levels of the self-healing execution library are:

[0071] C1, the basic layer, includes basic self-healing operations at the device level such as device restart, parameter reset, standby device switching, etc.

[0072] C2, the environmental layer, includes self-healing strategies for temperature and humidity, air conditioning, and fresh air environment control equipment;

[0073] C3, the network layer, includes self-healing solutions related to network communication and data transmission;

[0074] C4, application layer: self-healing measures for software systems and data processing, and each policy item includes: fault type identification, fault severity rating (1-5 levels), self-healing operation step sequence, operation timeout limit, expected effect indicator and historical execution success rate.

[0075] By setting up dynamic risk level assessment, building a multi-level risk assessment model, calculating the risk index of each component of the system in real time, analyzing the correlation risk between devices based on the Bayesian network, establishing a risk propagation model, assessing the possibility of chain reactions of failures, dynamically generating risk maps, and intuitively displaying the weak links of the system

[0076] Wherein, the self-healing execution unit includes a strategy matching layer, an execution control layer and an emergency processing layer;

[0077] By building a self-healing execution unit, it has the ability to self-repair and can solve certain common problems without human intervention.

[0078] The strategy matching layer will take into account multi-dimensional factors such as fault type, severity, historical success rate, etc., and retrieve the self-healing plan with the highest matching degree from the strategy library based on the fault diagnosis results, and then handle non-complete matching situations through fuzzy matching algorithm; the execution control layer will establish an operation queue, execute self-healing steps in order of priority, and set up an operation timeout mechanism to prevent deadlock. The execution control layer will monitor the execution status and results of each step in real time, and keep operation logs to record the detailed execution process; the emergency processing layer is equipped with a rollback mechanism to restore the system to its original state when self-healing fails, and when the self-healing failure reaches a preset number of times, the alarm level is automatically upgraded. The emergency processing layer will start backup equipment or switch redundant systems when necessary.

[0079] The self-healing execution unit also includes a self-healing knowledge base maintenance layer, which includes a data accumulation function: it will record detailed information of all self-healing operations, establish a fault-solution mapping relationship, and statistically analyze the laws of fault occurrence; an experience learning function: it uses machine learning algorithms to analyze historical data, extract effective self-healing patterns, and further generate new self-healing strategy recommendations; a continuous optimization function: it regularly evaluates the effectiveness of the strategy library, cleans up inefficient or outdated strategies, and introduces new self-healing methods and technologies.

[0080] By setting up active prevention and control strategies, a series of system operation parameters are automatically adjusted according to risk assessment results, intelligent balancing of equipment loads is achieved, and the pressure on high-risk components is reduced. When potential risks are predicted, preventive maintenance procedures are initiated in advance, and an emergency resource pre-positioning mechanism is established to ensure rapid response at critical moments.

[0081] The self-healing effect evaluation unit includes a real-time evaluation function, an effect analysis function and a strategy optimization function.

[0082] Among them, the real-time evaluation function will monitor the changes in key performance indicators (KPIs), and compare the system status before and after the failure to verify whether the expected repair effect is achieved; the effect analysis function will calculate the success rate of the self-healing operation and calculate the mean time to repair (MTTR). When the repair fails, the cause of the failure will be analyzed and a report will be generated; the policy optimization function will update the policy library according to the evaluation results, adjust the policy priority and self-healing parameters, and thus establish a policy version control mechanism.

[0083] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0084] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for risk assessment, fault diagnosis and fault self-healing based on a CEMS system, characterized by: This method realizes fault diagnosis and self-healing based on machine learning mechanism. The specific steps are as follows: Step 1: Environmental monitoring and data upload: deploy sensors for monitoring environmental parameters and upload environmental data to the host; Step 2: Remote control and feedback: Build a remote control and feedback system to allow technicians to remotely control the facilities and equipment in the computer room through a central monitoring platform; Step 3: Machine learning model training: Process the large amount of historical operation data collected in the CEMS system and train the model with the processed data so that it can automatically identify potential failure modes. Step 4: Establish a warning module. Construct a warning module so that it can detect potential failure risks in advance and issue warning signals through real-time monitoring and analysis of the CEMS system; Step 5: The self-healing module is established. When the model detects a fault in the system, the self-healing module will start the corresponding self-healing program to repair and adjust the system.

2. The method for risk assessment, fault diagnosis and fault self-healing based on cems system according to claim 1 is characterized by: The machine learning model can analyze system operation data in real time, automatically identify potential failure modes, and improve the accuracy of fault diagnosis through multi-parameter fusion analysis.

3. The method for risk assessment, fault diagnosis and fault self-healing based on CEMS system according to claim 1 is characterized by: The training method of the machine learning model is: A1, collect a large amount of historical operation data from the CEMS system and pre-process the data; A2, extract useful feature information from the preprocessed data, and this feature information will be used as the input vector of the machine learning model; A3, selecting a suitable machine learning model, and the machine learning model is one of support vector machine (SVM), random forest (RF) and neural network (NN); A4, train the selected model using the training data set and optimize the model parameters through cross-validation method; A5, inputs the real-time collected data into the trained machine learning model and outputs the fault diagnosis results.

4. The method for risk assessment, fault diagnosis and fault self-healing based on CEMS system according to claim 1 is characterized by: During the environmental monitoring and data uploading stage, the sensors deployed in the computer room upload the data to the host quickly and securely through the protocol, and the protocol must support breakpoint resumption and data encryption functions to ensure the integrity and security of the data.

5. The method for risk assessment, fault diagnosis and fault self-healing based on cems system according to claim 1 is characterized by: The remote control and feedback system allows technicians to remotely control the facilities and equipment in the computer room through the central monitoring platform, and provide real-time feedback on the operating status and operation results of the equipment.

6. The method for risk assessment, fault diagnosis and fault self-healing based on cems system according to claim 1 is characterized by: The steps of realizing the early warning function of the pre-warning module are as follows: B1, first analyze the historical data, identify the long-term change trend and abnormal points of the data, and combine the trend analysis results with the current data to evaluate the potential failure risk level; B2, when the assessment is completed and the assessment results show that there is a high risk of failure, an early warning signal is generated and notified to relevant personnel through interface display and email notification; B3, the early warning module will provide corresponding preventive measures based on different types of potential failure risks, helping users take timely measures to reduce the possibility of failure.

7. The method for risk assessment, fault diagnosis and fault self-healing based on cems system according to claim 1 is characterized by: The self-healing module includes a self-healing strategy library, a self-healing execution unit and a self-healing effect evaluation unit; The specific levels of the self-healing execution library are: C1, the basic layer, contains basic self-healing operations at the device level; C2, the environmental layer, includes self-healing strategies for temperature and humidity, air conditioning, and fresh air environment control equipment; C3, the network layer, includes self-healing solutions related to network communication and data transmission; C4, application layer: self-healing measures for software systems and data processing, and each policy item includes: fault type identification, fault severity rating, self-healing operation step sequence, operation timeout limit, expected effect indicator and historical execution success rate.

8. The method for risk assessment, fault diagnosis and fault self-healing based on cems system according to claim 7 is characterized by: The self-healing execution unit includes a strategy matching layer, an execution control layer and an emergency processing layer; The strategy matching layer will take multi-dimensional factors into consideration and retrieve the self-healing solution with the highest matching degree from the strategy library based on the fault diagnosis results; the execution control layer will establish an operation queue, execute the self-healing steps in order of priority, and set up an operation timeout mechanism to prevent deadlock; the emergency processing layer will set up a rollback mechanism to restore the system to its original state when self-healing fails, and automatically upgrade the alarm level when the self-healing failure reaches a preset number of times.

9. The method for risk assessment, fault diagnosis and fault self-healing based on cems system according to claim 1, characterized in that: The self-healing effect evaluation unit includes a real-time evaluation function, an effect analysis function and a strategy optimization function.

10. The method for risk assessment, fault diagnosis and fault self-healing based on CEMS system according to claim 9, characterized in that: The real-time evaluation function will monitor changes in key performance indicators and compare the system status before and after the failure to verify whether the expected repair effect is achieved; the effect analysis function will calculate the success rate of the self-healing operation and calculate the average repair time. When the repair fails, it will analyze the cause of the failure and form a report; the policy optimization function will update the policy library according to the evaluation results and adjust the policy priority and self-healing parameters.

Images