CN-122020339-A - Intelligent ammeter full life cycle fault root cause analysis method and system

CN122020339ACN 122020339 ACN122020339 ACN 122020339ACN-122020339-A

Abstract

The invention belongs to the technical field of computing devices based on knowledge models, and particularly relates to a full life cycle fault root cause analysis method and system of an intelligent electric energy meter, wherein the method comprises the steps of obtaining a multi-mode operation data set of the intelligent electric energy meter; the method comprises the steps of determining initial nuclear weight adjusting factors, equipment aging compensation factors, optimized nuclear weight adjusting factors and weights of all nuclear functions after optimization of the intelligent electric energy meter, reconstructing target discriminant functions of a multi-core support vector machine of the intelligent electric energy meter, classifying fault types of the intelligent electric energy meter, tracking returning time nodes of heat mutation generated by the intelligent electric energy meter, and locking fault root causes. According to the invention, by analyzing the numerical characteristics of the multi-mode operation data of the intelligent electric energy meter, a dynamic kernel weight adjusting mechanism is constructed, so that the classification accuracy of an algorithm is enhanced, and the productivity benefit and the service life of the intelligent electric energy meter are improved.

Inventors

ZHANG JIE
WANG GANG
ZHANG JIANCHUN

Assignees

江阴众和电力仪表有限公司

Dates

Publication Date: 20260512
Application Date: 20260415

Claims (10)

1. The full life cycle fault root cause analysis method of the intelligent electric energy meter is characterized by comprising the following steps of: Collecting full life cycle circulation record data of the intelligent electric energy meter, and preprocessing to obtain a multi-mode operation data set of the intelligent electric energy meter; Determining an initial nuclear weight adjustment coefficient of the intelligent electric energy meter according to the difference value between the workshop environment temperature in the multi-mode dataset and the standard workshop reference temperature; Determining an equipment aging compensation factor of the intelligent electric energy meter according to the difference value between the actual running output power average value of each heating equipment in the multi-mode data set and the rated standard power of each heating equipment in the factory healthy state, and determining an optimized core weight adjustment factor of the intelligent electric energy meter by combining the initial core weight adjustment factor and the equipment aging compensation factor; Determining the weight of each core function of the intelligent electric energy meter after optimization by using the optimization core weight adjustment coefficient, and reconstructing a target discriminant function of a multi-core support vector machine of the intelligent electric energy meter; And classifying the fault types of the intelligent electric energy meter by using the target discrimination function, judging the difference value between the actual operation output power of each heating device and the rated standard power by using a preset danger judgment threshold value based on the classification result, tracking the return time node of the heat mutation generated by the intelligent electric energy meter, and locking the root cause of the fault.
2. The method for analyzing the root cause of the full life cycle fault of the intelligent ammeter according to claim 1, wherein the method for acquiring the initial kernel weight adjustment coefficient is characterized in that an absolute value of a difference value between the workshop environment temperature and a standard workshop reference temperature in a multi-mode dataset is calculated to obtain a first value, a ratio between the first value and the standard workshop reference temperature is calculated to obtain a second value, a natural exponential function is applied to the second value, and the obtained result is used as the initial kernel weight adjustment coefficient.
3. The method for analyzing the root cause of the full life cycle fault of the intelligent ammeter according to claim 1 is characterized in that the method for acquiring the equipment aging compensation factor is that a difference value between an average value of actual running output power of each heating equipment in a multi-mode data set and rated standard power of the heating equipment in a factory healthy state is calculated to obtain a third value, a ratio of the third value to the rated standard power of each heating equipment in the factory healthy state is calculated to obtain a fourth value, the fourth values corresponding to all the heating equipment are summed, and the obtained result is used as the equipment aging compensation factor.
4. The method for analyzing the root cause of the full life cycle fault of the intelligent electric energy meter according to claim 1, wherein the method for obtaining the optimized core weight adjusting coefficient is characterized by comprising the steps of applying a natural exponential function to the equipment aging compensation factor to obtain a fifth value, calculating the product of the fifth value and the equipment aging compensation factor, calculating the sum of the product and the initial core weight adjusting coefficient, and taking the obtained result as the optimized core weight adjusting coefficient.
5. The method for analyzing the root cause of the full life cycle fault of the intelligent electric energy meter according to claim 1, wherein the method for obtaining the optimized weight of each kernel function is characterized in that a natural exponential function is applied to the optimized kernel weight adjusting coefficient, the optimized kernel weight adjusting coefficient after the natural exponential function is used as an exponential term of an initial weight of each kernel function, the obtained result is used as a sixth numerical value, the ratio between the sixth numerical value of each kernel function and the sum of the sixth numerical values of all kernel functions is calculated, and the obtained result is used as the optimized weight of each kernel function.
6. The method for analyzing the root cause of the full life cycle fault of the intelligent electric energy meter according to claim 1, wherein the reconstructing the objective discriminant function of the multi-core support vector machine of the intelligent electric energy meter comprises substituting the optimized weights of the kernel functions into an initial objective discriminant function, and replacing the initial weights by the optimized weights of the kernel functions of the intelligent electric energy meter to complete the reconstruction of the objective discriminant function of the multi-core support vector machine of the intelligent electric energy meter.
7. The method for analyzing the full life cycle fault root cause of the intelligent electric energy meter according to claim 1, wherein the method for tracking the time node of the return of the heat mutation generated by the intelligent electric energy meter for locking the fault root cause comprises the steps of carrying out reverse path tracking according to a kernel function corresponding to actual operation output power data of heating equipment corresponding to all secondary reworking procedures of the intelligent electric energy meter in response to a classification result being malignant heat damage, and locking the fault root cause in response to the fact that the difference value between the actual operation output power data of the heating equipment corresponding to the primary reworking procedure and the rated standard power is larger than a preset dangerous judgment threshold value, wherein the time node where the actual operation output power data of the heating equipment corresponding to the secondary reworking procedure is the time node of the return of the heat mutation generated by the intelligent electric energy meter.
8. The method for analyzing the root cause of the full life cycle fault of the intelligent electric energy meter according to claim 1, wherein the step of collecting the full life cycle circulation record data of the intelligent electric energy meter comprises the step of collecting the full life cycle circulation record data of the intelligent electric energy meter through an internet of things sensor cluster deployed in each production workshop, wherein the full life cycle circulation record data comprises text data of all reworking procedures of the intelligent electric energy meter, actual operation output power data of heating equipment corresponding to all reworking procedures of the intelligent electric energy meter, real-time heating temperature data and workshop environment temperature data in the production period of the intelligent electric energy meter.
9. The full life cycle fault cause analysis method of the intelligent electric energy meter according to claim 1, wherein the preprocessing is performed to obtain a multi-mode operation data set of the intelligent electric energy meter, and the method comprises the steps of performing time stamp alignment operation on collected text data of all reworking procedure sequences of the intelligent electric energy meter, actual operation output power data of heating equipment corresponding to all reworking procedures of the intelligent electric energy meter, real-time heating temperature data and workshop environment temperature data in a production period of the intelligent electric energy meter based on unique medium access control address identification codes burnt by the intelligent electric energy meter, and constructing the multi-mode operation data set of the intelligent electric energy meter.
10. The full life cycle fault root cause analysis system of the intelligent electric energy meter is characterized by comprising a processor and a memory, wherein the memory stores computer program instructions, and the computer program instructions are executed by the processor to realize the full life cycle fault root cause analysis method of the intelligent electric energy meter according to any one of claims 1-9.

Description

Intelligent ammeter full life cycle fault root cause analysis method and system Technical Field The invention relates to the technical field of computing devices based on knowledge models. More particularly, the invention relates to a full life cycle fault root analysis method and system for an intelligent electric energy meter. Background Along with the continuous deepening and popularization of smart power grid construction, the intelligent electric energy meter is used as a core terminal device for data acquisition and metering, the long-term operation reliability of the intelligent electric energy meter under a complex field environment is very important, as the production and manufacture of the intelligent electric energy meter involve a plurality of complex and precise automatic full life cycle processes such as surface mounting, dual in-line packaging, wave soldering and the like, a printed circuit board often undergoes complex reworking flows such as repeated manual repair welding and secondary furnace passing and the like due to misjudgment of an automatic instrument or substandard test parameters in the actual pipeline production process, and the irregular repeated reworking can lead tiny electronic components on the printed circuit board to bear repeated nonstandard local thermal shocks so as to form hidden thermal stress damage, and the damage is normally represented in factory testing, but is extremely easy to cause serious faults such as pad stripping, off-line of a communication module and the like after long-term full-load operation. At present, a multi-core support vector machine algorithm is often utilized to mine the association of discrete reworked texts with continuous temperature and power curves so as to conduct classification judgment, however, after model training is completed, the weight distribution proportion among all kernel functions in the multi-core support vector machine algorithm is in a static fixed state, the macroscopic environment reference temperature of a workshop in an actual automatic production line can generate extremely remarkable drifting phenomenon along with season replacement, for example, extremely cold environments in winter are extremely fast in heat dissipation, high-temperature environment heat is extremely easy to generate long-term accumulation effect in deep base materials in summer, static kernel weights are completely difficult to perceive and adapt to macroscopic seasonal drifting, decision hyperplane rigidity of the static kernel weights is caused, the benign reworked curves which are safe in winter possibly evolve into malignant data which cause thermal stress damage in summer, and meanwhile, a static curing mechanism is more incapable of detecting fatal damage caused by deep hidden thermal energy superposition when the underlying power surge phenomenon is caused by microscopic aging of underlying heating equipment, so that the algorithm is easy to generate serious misjudgment when the algorithm is used for processing the historical running data of crossing seasons and month lots, and the complete life cycle of intelligent electric energy meter is not capable of being accurately traced back from a real life cycle root of a complicated production network. Disclosure of Invention In order to solve the technical problems that the weight of each kernel function of the existing multi-kernel support vector machine algorithm is statically fixed, the existing multi-kernel support vector machine algorithm is difficult to adapt to seasonal temperature drift of an intelligent electric energy meter production workshop and ageing environments of heating equipment, fault type classification misjudgment is caused, and heat mutation causes cannot be accurately traced, the invention provides a scheme in the following aspects. The invention provides a full life cycle fault cause analysis method of an intelligent electric energy meter, which comprises the steps of collecting full life cycle circulation record data of the intelligent electric energy meter, preprocessing to obtain a multi-mode operation data set of the intelligent electric energy meter, determining initial nuclear weight adjustment coefficients of the intelligent electric energy meter according to a difference value between workshop environment temperature and standard workshop reference temperature in the multi-mode data set, determining equipment aging compensation factors of the intelligent electric energy meter according to a difference value between an actual operation output power average value of each heating equipment in the multi-mode data set and rated standard power of the heating equipment in a factory healthy state, determining optimal nuclear weight adjustment coefficients of the intelligent electric energy meter by combining the initial nuclear weight adjustment coefficients and the equipment aging compensation factors, determining weights of the intelligent electric energy meter after