WO2023241042A1 - 故障预测方法、装置、电子设备及存储介质 - Google Patents

故障预测方法、装置、电子设备及存储介质 Download PDF

Info

Publication number
WO2023241042A1
WO2023241042A1 PCT/CN2023/072164 CN2023072164W WO2023241042A1 WO 2023241042 A1 WO2023241042 A1 WO 2023241042A1 CN 2023072164 W CN2023072164 W CN 2023072164W WO 2023241042 A1 WO2023241042 A1 WO 2023241042A1
Authority
WO
WIPO (PCT)
Prior art keywords
network management
prediction model
fault
model
management system
Prior art date
Application number
PCT/CN2023/072164
Other languages
English (en)
French (fr)
Inventor
陈洋
Original Assignee
中兴通讯股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Publication of WO2023241042A1 publication Critical patent/WO2023241042A1/zh

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/06—Management of faults, events, alarms or notifications
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H04L41/147—Network analysis or design for predicting network behaviour
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H04L41/149—Network analysis or design for prediction of maintenance
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W24/00—Supervisory, monitoring or testing arrangements
    • H04W24/04—Arrangements for maintaining operational condition
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W24/00—Supervisory, monitoring or testing arrangements
    • H04W24/06—Testing, supervising or monitoring using simulated traffic

Definitions

  • the present disclosure relates to the technical field of communication equipment, and in particular to a fault prediction method, device, electronic equipment and storage medium.
  • telecommunications base stations are generally equipped with BBU (Building Base Band Unit, indoor baseband processing unit) equipment and RRU (Remote Radio Unit, radio frequency remote module) equipment, and the two are connected with optical fibers to solve the indoor network of large venues. cover.
  • BBU Building Base Band Unit, indoor baseband processing unit
  • RRU Remote Radio Unit, radio frequency remote module
  • Embodiments of the present disclosure provide a fault prediction method, device, electronic equipment, and storage medium.
  • a fault prediction method includes: obtaining fault prediction models uploaded by multiple network management systems, wherein each network management system performs model training based on base station sample data corresponding to the network management system to obtain the A fault prediction model corresponding to the network management system; determining at least one candidate prediction model among a plurality of the fault prediction models, and determining a target prediction model based on at least one of the candidate prediction models; and issuing the target prediction model to a plurality of all
  • the network management system is configured to enable multiple network management systems to use the target prediction model to perform fault prediction on corresponding base stations.
  • a fault prediction device in a second aspect, includes: an acquisition module configured to acquire fault prediction models uploaded by multiple network management systems, wherein each network management system is based on base station sample data corresponding to the network management system. Perform model training to obtain a fault prediction model corresponding to the network management system; a determination module configured to determine at least one candidate prediction model among a plurality of the fault prediction models, and determine a target prediction model based on at least one of the candidate prediction models; and a delivery module configured to deliver the target prediction model to multiple network management systems, so that multiple network management systems use the target prediction model to perform fault prediction on corresponding base stations.
  • an electronic device including a processor, a communication interface, a memory, and a communication bus.
  • the processor, the communication interface, and the memory complete communication with each other through the communication bus;
  • the memory is used to store computer programs;
  • the processor is used to implement the method described in the first aspect when executing the program stored in the memory.
  • a computer-readable storage medium is provided.
  • a computer program is stored in the computer-readable storage medium.
  • the computer program is executed by a processor, the method described in the first aspect is implemented.
  • a computer program product containing instructions which when run on a computer causes the computer to execute the above fault prediction method.
  • Figure 1 is a schematic diagram of an application scenario of a fault prediction method provided by an embodiment of the present disclosure
  • Figure 2 is a flow chart of a fault prediction method provided by an embodiment of the present disclosure
  • Figure 3 is a flow chart of another fault prediction method provided by an embodiment of the present disclosure.
  • Figure 4 is a schematic structural diagram of a fault prediction device provided by an embodiment of the present disclosure.
  • FIG. 5 is a schematic structural diagram of an electronic device provided by an embodiment of the present disclosure.
  • the BBU equipment may overheat. If the temperature environment health of the BBU equipment is very poor, it may cause business interruption due to power outage of the equipment or equipment damage and repair. When it is discovered that the business is interrupted and the station is inspected for maintenance, the damage and impact caused are irreparable. Therefore, how to predict the overheating of base station BBU equipment in advance has become an urgent problem to be solved.
  • Figure 1 shows a schematic diagram of an application scenario of a fault prediction method provided by the present disclosure, in which a server is communicatively connected to multiple network management systems, and each network management system is communicatively connected to its corresponding base station through a gateway.
  • This allows the server to use federated learning to aggregate local data samples from different network management systems through local models trained by local machine learning to form a final joint model, and then send the joint model to the network management system for combination
  • a fault prediction method provided by this disclosure can be applied to servers to predict faults on base station equipment, such as predicting over-temperature of BBU (Building Base Band Unit, indoor baseband processing unit) equipment.
  • BBU Building Base Band Unit, indoor baseband processing unit
  • Each network management system collects base station sample data used to train fault prediction models from the corresponding base stations, and locally uses machine learning to train local fault prediction models (such as BBU equipment over-temperature prediction models). Then, upload the trained fault prediction model to the server.
  • local fault prediction models such as BBU equipment over-temperature prediction models
  • the fault prediction model uploaded by multiple network management systems can be obtained in the following manner: for any network management system among the multiple network management systems, obtain the first encrypted data uploaded by the network management system, and determine the first encrypted data corresponding to the network management system. Rules, wherein the network management system encrypts the corresponding fault prediction model based on the first encryption rule to obtain the first encrypted data; and decrypts the first encrypted data based on the first encryption rule to obtain the corresponding fault prediction model of the network management system.
  • the first encryption rule corresponding to each network management system can be set in advance.
  • the network management system uploads the first encrypted data corresponding to the fault prediction model
  • the first encrypted data can be decrypted according to the first encryption rule corresponding to the network management system to obtain Failure prediction model. Thereby improving the security of data transmission.
  • S102 Determine at least one candidate prediction model among multiple fault prediction models, and determine a target prediction model based on at least one candidate prediction model.
  • fault prediction models uploaded by multiple network management systems can be screened to filter out fault prediction models with lower accuracy to obtain at least one candidate prediction model with higher accuracy, and then at least one candidate prediction model can be obtained.
  • the prediction models are aggregated to obtain a joint model, which is the target prediction model.
  • At least one candidate prediction model can be determined among multiple fault prediction models in the following manner: for any fault prediction model among the multiple fault prediction models, obtain the number of samples corresponding to the fault prediction model; and convert the number of samples Fault prediction models that are greater than or equal to the preset quantity threshold are determined as candidate prediction models.
  • each network management system When each network management system uploads a fault prediction model, it can also upload the total amount of sample data for training the fault prediction model, that is, the number of samples. If the number of samples is less than the preset number threshold, it means that the accuracy of the model is low and it does not participate in the aggregation calculation of the target prediction model; if the number of samples is greater than or equal to the preset number threshold, it means that the accuracy of the model is high and it does not participate in the aggregation calculation of the target prediction model. Aggregation calculation of the model, thereby improving the accuracy of the target prediction model.
  • the target prediction model can be delivered to multiple network management systems, and the network management system uses the received target prediction model to replace the local fault prediction model to perform fault prediction on the base station equipment.
  • the target prediction model can be delivered to multiple network management systems in the following manner: encrypt the target prediction model based on the second encryption rule to obtain second encrypted data; deliver the second encrypted data to multiple network management systems, In this way, multiple network management systems decrypt the second encrypted data based on the second encryption rule to obtain the target prediction model.
  • the second encryption rule can be preset in the server, and the target prediction model is encrypted using the second encryption rule and then distributed to multiple network management systems, thereby improving the security of data transmission.
  • fault prediction models uploaded by multiple network management systems are obtained, in which each network management system performs model training based on base station sample data corresponding to the network management system to obtain a fault prediction model corresponding to the network management system; then, in multiple network management systems, fault prediction models are obtained. Determine at least one candidate prediction model in the fault prediction model, and determine the target prediction model based on the at least one candidate prediction model; finally, deliver the target prediction model to multiple network management systems, so that the multiple network management systems use the target prediction model to predict the corresponding base station Make failure predictions.
  • base station equipment failures can be predicted, fault problems can be discovered in advance, losses can be reduced, and users can be provided with more stable communication services.
  • local models i.e., candidate prediction models
  • target prediction model i.e., target prediction model
  • the network management system can train the fault prediction model in the following manner.
  • feature function (1) is as follows:
  • x1...xn represents the characteristic variables represented by a to i in step (1)
  • y is the expected result
  • c1...cn represents the characteristic parameters of the characteristic variables represented by a to i
  • c0 is a constant parameter.
  • the model is then continuously updated using gradient descent.
  • the loss function fluctuates very little with the gradient, it means that it has basically converged to the minimum value.
  • the model training is completed.
  • c0, c1, c2...cn are the parameters of the trained fault prediction model.
  • S102 may include the following steps S201 to S203.
  • each candidate prediction model For each candidate prediction model, obtain the weight corresponding to the candidate prediction model, and obtain the model parameters of the candidate prediction model, where at least one candidate prediction model is trained based on the same preset algorithm.
  • S202 Perform a weighted average operation based on the model parameters and weights corresponding to at least one candidate prediction model to obtain the target parameters.
  • S203 use the target parameters to configure the preset algorithm to obtain the target prediction model.
  • the candidate prediction model is trained based on the same preset algorithm (such as feature function (1)). You can set the weight of each candidate prediction model, and use this weight to perform a weighted average operation on the model parameters corresponding to all candidate prediction models to obtain the target parameters. When there are multiple model parameters, a weighted average is performed for each model parameter. Operation to obtain the corresponding target parameters. Then use the target parameters to configure the preset algorithm to obtain the target prediction model.
  • the same preset algorithm such as feature function (1)
  • the preset algorithm includes parameters a and b
  • the candidate prediction models include model one and model two.
  • obtaining the weight corresponding to the candidate prediction model may include the following steps: obtaining the number of fault samples corresponding to the candidate prediction model, and the total number of fault samples corresponding to at least one candidate prediction model; calculating the number of fault samples and the total number of fault samples The ratio of the quantities is used as the corresponding weight of the candidate prediction model.
  • each network management system When each network management system uploads a fault prediction model, it can also upload the number of fault samples for training the fault prediction model (that is, the number of risk and fault state samples). The number of fault samples of all candidate prediction models is summed to obtain all The total number of fault samples for candidate prediction models. Furthermore, for each candidate prediction model, the ratio of the corresponding number of fault samples to the total number of fault samples is used as the weight corresponding to the candidate prediction model.
  • network management system T1 M1/(M1+M2+M3)
  • the weights corresponding to the candidate prediction models can be used to perform a weighted average operation on the model parameters to obtain the target parameters, and then the target parameter configuration preset algorithm can be used to obtain the target prediction model. Thereby improving the prediction accuracy of the target prediction model.
  • the weight of each candidate prediction model is the same, and then the target parameters can be obtained by directly calculating the average value of the model parameters, thereby reducing the calculation difficulty.
  • the method may further include the following steps: receiving prediction results sent by multiple network management systems, wherein each network management system uses a target prediction model to perform fault prediction on the corresponding base station to obtain prediction results corresponding to the base station. ; and generate a base station fault report based on multiple prediction results.
  • each network management system can input the characteristic data collected in real time into the target prediction model to obtain corresponding prediction results.
  • the prediction results generally include three situations: normal, risk and failure.
  • the prediction results are uploaded to the server, and the server can generate a base station fault report based on the prediction results uploaded by multiple network management systems, where the base station fault report includes the base station identification and prediction results of each base station.
  • the base station fault report may also include corresponding fault problems and processing methods. This makes it easier for staff to check the health status of multiple base stations.
  • the device includes: an acquisition module 301 configured to acquire fault prediction models uploaded by multiple network management systems, wherein each A network management system performs model training based on base station sample data corresponding to the network management system to obtain a fault prediction model corresponding to the network management system; the determination module 302 is configured to determine at least one candidate prediction model among multiple fault prediction models, and based on at least one candidate prediction The model determines the target prediction model; and the delivery module 303 is configured to deliver the target prediction model to multiple network management systems, so that the multiple network management systems use the target prediction model to perform fault prediction on the corresponding base station.
  • the determination module is configured to: for each candidate prediction model, obtain the weight corresponding to the candidate prediction model, and obtain the model parameters of the candidate prediction model, wherein at least one candidate prediction model is based on the same A preset algorithm is trained; a weighted average operation is performed based on the model parameters and weights corresponding to at least one candidate prediction model to obtain the target parameters; and the preset algorithm is configured using the target parameters to obtain the target prediction model.
  • the determination module is further configured to: obtain the number of fault samples corresponding to the candidate prediction models, and the total number of fault samples corresponding to at least one candidate prediction model; and calculate the number of fault samples and the total number of fault samples The ratio is used as the corresponding weight of the candidate prediction model.
  • the determination module is further configured to: obtain the number of samples corresponding to the fault prediction model for any one of the multiple fault prediction models; and set the number of samples to be greater than or equal to a preset number threshold.
  • the fault prediction model is determined as a candidate prediction model.
  • the acquisition module is configured to: obtain the first encrypted data uploaded by the network management system for any one of the multiple network management systems, and determine the first encryption rule corresponding to the network management system, where , the network management system encrypts the corresponding fault prediction model based on the first encryption rule to obtain the first encrypted data; and decrypts the first encrypted data based on the first encryption rule to obtain the corresponding fault prediction model of the network management system.
  • the delivery module is configured to: encrypt the target prediction model based on the second encryption rule to obtain second encrypted data; and deliver the second encrypted data to multiple network management systems so that multiple A network management system decrypts the second encrypted data based on the second encryption rule to obtain a target prediction model.
  • the device further includes a generation module configured to: receive prediction results sent by multiple network management systems, wherein each network management system uses the target prediction model to perform fault prediction on the corresponding base station to obtain the prediction corresponding to the base station. results; and based on multiple prediction results, generate a base station fault report.
  • a generation module configured to: receive prediction results sent by multiple network management systems, wherein each network management system uses the target prediction model to perform fault prediction on the corresponding base station to obtain the prediction corresponding to the base station. results; and based on multiple prediction results, generate a base station fault report.
  • fault prediction models uploaded by multiple network management systems are obtained, in which each network management system performs model training based on base station sample data corresponding to the network management system to obtain a fault prediction model corresponding to the network management system; then, in multiple network management systems, fault prediction models are obtained. Determine at least one candidate prediction model in the fault prediction model, and determine the target prediction model based on the at least one candidate prediction model; finally, deliver the target prediction model to multiple network management systems, so that the multiple network management systems use the target prediction model to predict the corresponding base station Make failure predictions.
  • base station equipment failures can be predicted, fault problems can be discovered in advance, losses can be reduced, and users can be provided with more stable communication services.
  • local models i.e., candidate prediction models
  • target prediction model i.e., target prediction model
  • embodiments of the present disclosure also provide an electronic device, as shown in FIG. 5 , including a processor 111, a communication interface 112, a memory 113, and a communication bus 114.
  • the processor 111, the communication interface 112, and the memory 113 complete communication with each other through the communication bus 114.
  • Memory 113 is used to store computer programs.
  • the processor 111 is used to execute the program stored on the memory 113 to implement the following steps: obtain fault prediction models uploaded by multiple network management systems, wherein each network management system performs model training based on the base station sample data corresponding to the network management system to obtain the network management system.
  • a fault prediction model corresponding to the system determining at least one candidate prediction model among multiple fault prediction models, and determining a target prediction model based on at least one candidate prediction model; and delivering the target prediction model to multiple network management systems so that multiple network management systems
  • the system uses the target prediction model to predict the failure of the corresponding base station.
  • the communication bus mentioned in the above-mentioned electronic equipment can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc.
  • PCI Peripheral Component Interconnect
  • EISA Extended Industry Standard Architecture
  • the communication bus can be divided into address bus, data bus, control bus, etc. For ease of presentation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.
  • the communication interface is used for communication between the above-mentioned electronic devices and other devices.
  • the memory may include Random Access Memory (RAM) or non-volatile memory.
  • RAM Random Access Memory
  • NVM Non-Volatile Memory
  • the memory may also be at least one storage device located far away from the aforementioned processor.
  • the above-mentioned processor can be a general-purpose processor, including a central processing unit (Central Processing Unit, CPU), a network processor (Network Processor, NP), etc.; it can also be a digital signal processor (Digital Signal Processing, DSP), special integrated Circuit (Application Specific Integrated Circuit, ASIC), Field-Programmable Gate Array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.
  • CPU Central Processing Unit
  • NP Network Processor
  • DSP Digital Signal Processing
  • ASIC Application Specific Integrated Circuit
  • FPGA Field-Programmable Gate Array
  • a computer-readable storage medium stores a computer program.
  • the steps of any of the above fault prediction methods are implemented. .
  • a computer program product containing instructions is also provided, which when run on a computer causes the computer to execute any of the fault prediction methods in the above embodiments.
  • a computer program product includes one or more computer instructions.
  • Computer instructions may be stored in or transmitted from one computer-readable storage medium to another computer-readable storage medium, e.g., computer instructions may be transmitted from a website, computer, server or data center via a wired link (e.g.
  • Coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless means to transmit to another website site, computer, server or data center.
  • Computer-readable storage media can be any available media that can be accessed by a computer or a data storage device such as a server, data center, or other integrated media that contains one or more available media. Available media may be magnetic media (eg, floppy disk, hard disk, tape), optical media (eg, DVD), or semiconductor media (eg, Solid State Disk (SSD)), etc.
  • Embodiments of the present disclosure provide a fault prediction method, device, electronic equipment and storage medium.
  • the present disclosure first obtains fault prediction models uploaded by multiple network management systems, wherein each network management system is based on base station samples corresponding to the network management system. Model training is performed on the data to obtain a fault prediction model corresponding to the network management system; then, at least one candidate prediction model is determined among multiple fault prediction models, and a target prediction model is determined based on at least one candidate prediction model; finally, the target prediction model is delivered to up to A network management system enables multiple network management systems to use target prediction models to predict faults for corresponding base stations. In this way, base station equipment failures can be predicted, fault problems can be discovered in advance, losses can be reduced, and users can be provided with more stable communication services.
  • a final joint model i.e., target prediction model
  • the joint model is used to predict base station failures. Avoid local data limitations affecting the accuracy of fault prediction.
  • implementing any product or method of the present disclosure does not necessarily require achieving all the above advantages simultaneously.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

本公开提供了一种故障预测方法、装置、电子设备及存储介质，属于通讯设备技术领域。本公开通过，获取多个网管系统上传的故障预测模型，其中，每个网管系统基于网管系统对应的基站样本数据进行模型训练得到网管系统对应的故障预测模型；在多个故障预测模型中确定至少一个候选预测模型，并基于至少一个候选预测模型确定目标预测模型；将目标预测模型下发至多个网管系统，以使多个网管系统利用目标预测模型对对应的基站进行故障预测。

Description

故障预测方法、装置、电子设备及存储介质
相关申请的交叉引用
本公开要求享有2022年06月13日提交的名称为“故障预测方法、装置、电子设备及存储介质”的中国专利申请CN202210666827.0的优先权，其全部内容通过引用并入本公开中。
技术领域
本公开涉及通讯设备技术领域，尤其涉及一种故障预测方法、装置、电子设备及存储介质。
背景技术
目前，电信基站处一般设有BBU(Building Base band Unite，室内基带处理单元)设备和RRU(Remote Radio Unit，射频拉远模块)设备，二者之间用光纤连接，以解决大型场馆的室内网络覆盖。
发明内容
本公开的实施例提供了一种故障预测方法、装置、电子设备及存储介质。
第一方面，提供了一种故障预测方法，所述方法包括：获取多个网管系统上传的故障预测模型，其中，每个网管系统基于所述网管系统对应的基站样本数据进行模型训练得到所述网管系统对应的故障预测模型；在多个所述故障预测模型中确定至少一个候选预测模型，并基于至少一个所述候选预测模型确定目标预测模型；以及将所述目标预测模型下发至多个所述网管系统，以使多个所述网管系统利用所述目标预测模型对对应的基站进行故障预测。
第二方面，提供了一种故障预测装置，所述装置包括：获取模块，被配置为获取多个网管系统上传的故障预测模型，其中，每个网管系统基于所述网管系统对应的基站样本数据进行模型训练得到所述网管系统对应的故障预测模型；确定模块，被配置为在多个所述故障预测模型中确定至少一个候选预测模型，并基于至少一个所述候选预测模型确定目标预测模型；以及下发模块，被配置为将所述目标预测模型下发至多个所述网管系统，以使多个所述网管系统利用所述目标预测模型对对应的基站进行故障预测。
第三方面，提供了一种电子设备，包括处理器、通信接口、存储器和通信总线，处理器、通信接口、存储器通过通信总线完成相互间的通信；存储器，用于存放计算机程序； 处理器，用于执行存储器上所存放的程序时，实现第一方面所述的方法。
第四方面，提供了一种计算机可读存储介质，所述计算机可读存储介质内存储有计算机程序，所述计算机程序被处理器执行时实现第一方面所述的方法。
第五方面，提供了一种包含指令的计算机程序产品，当其在计算机上运行时，使得计算机执行上述的故障预测方法。
附图说明
为了更清楚地说明本公开实施例或相关技术中的技术方案，下面将对实施例或相关技术描述中所需要使用的附图作简单地介绍，显而易见地，对于本领域普通技术人员而言，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。
图1为本公开实施例提供的一种故障预测方法的应用场景示意图；
图2为本公开实施例提供的一种故障预测方法的流程图；
图3为本公开实施例提供的另一种故障预测方法的流程图；
图4为本公开实施例提供的一种故障预测装置的结构示意图；以及
图5为本公开实施例提供的一种电子设备的结构示意图。
具体实施方式
下面将结合本公开实施例中的附图，对本公开实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本公开一部分实施例，而不是全部的实施例。基于本公开中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本公开保护的范围。
针对基站BBU设备而言，由于单板故障或者安装等问题会导致BBU设备过温，如果BBU设备温度环境健康度很差，会导致设备下电引起的业务中断或者设备损坏返修。当发现业务中断再去上站检修，造成的损失和影响已经无法挽回。因此，如何提前预测基站BBU设备过温成为亟待解决的问题。
图1所示为本公开提供的一种故障预测方法的应用场景示意图，其中，服务器与多个网管系统通信连接，每个网管系统通过网关与其对应的基站通信连接。从而使服务器可以采用联邦学习的方式，对不同网管系统利用本地的数据样本通过本地机器学习训练的本地模型进行聚合，形成最终的联合模型，并将联合模型下发至网管系统，由网管系统结合现网收集的故障特征数据，来预测基站故障，避免基站本地样本特征数据的局限性导致预测模型失真。
为便于对本公开实施例的理解，下面将结合具体实施方式，对本公开实施例提供的一 种故障预测方法进行详细的说明，如图2所示，包括步骤S101至步骤S103。
S101，获取多个网管系统上传的故障预测模型，其中，每个网管系统基于网管系统对应的基站样本数据进行模型训练得到网管系统对应的故障预测模型。
本公开提供的一种故障预测方法可以应用于服务器，用于对基站设备进行故障预测，如对BBU(Building Base band Unite，室内基带处理单元)设备过温进行预测。
每个网管系统从对应的基站上采集用于训练故障预测模型的基站样本数据，并在本地利用机器学习训练本地的故障预测模型(如BBU设备过温预测模型)。进而，将训练好的故障预测模型上传至服务器。
作为实施方式，可以通过以下方式获取多个网管系统上传的故障预测模型：针对多个网管系统中的任一网管系统，获取网管系统上传的第一加密数据，以及确定网管系统对应的第一加密规则，其中，网管系统基于第一加密规则将对应的故障预测模型进行加密得到第一加密数据；基于第一加密规则对第一加密数据进行解密，得到网管系统对应的故障预测模型。
可以预先设置每个网管系统对应的第一加密规则，当网管系统上传故障预测模型对应的第一加密数据时，则可以按照该网管系统对应的第一加密规则对第一加密数据进行解密，得到故障预测模型。从而提高数据传输的安全性。
S102，在多个故障预测模型中确定至少一个候选预测模型，并基于至少一个候选预测模型确定目标预测模型。
在本公开实施例中，可以在多个网管系统上传的故障预测模型中进行筛选，筛选掉准确性较低的故障预测模型，得到准确性较高的至少一个候选预测模型，进而对至少一个候选预测模型进行聚合得到联合模型，即目标预测模型。
作为一些实施方式，可以通过以下方式在多个故障预测模型中确定至少一个候选预测模型：针对多个故障预测模型中的任一故障预测模型，获取故障预测模型对应的样本数量；以及将样本数量大于或等于预设数量阈值的故障预测模型，确定为候选预测模型。
每个网管系统在上传故障预测模型时，可以同时上传训练该故障预测模型的样本数据总量，即样本数量。样本数量小于预设数量阈值，意味着该模型的准确性较低，不参与目标预测模型的聚合计算；样本数量大于或等于预设数量阈值，意味着该模型的准确性较高，参与目标预测模型的聚合计算，从而提高目标预测模型的准确性。
S103，将目标预测模型下发至多个网管系统，以使多个网管系统利用目标预测模型对对应的基站进行故障预测。
在本公开实施例中，可以将目标预测模型下发至多个网管系统，网管系统利用接收到的目标预测模型替换本地的故障预测模型，对基站设备进行故障预测。
作为一些实施方式，可以通过以下方式将目标预测模型下发至多个网管系统：基于第二加密规则对目标预测模型进行加密，得到第二加密数据；将第二加密数据下发至多个网管系统，以使多个网管系统基于第二加密规则对第二加密数据进行解密得到目标预测模型。
服务器中可以预先设置第二加密规则，利用第二加密规则对目标预测模型进行加密后，下发至多个网管系统，从而提高数据传输的安全性。
本公开实施例中，首先，获取多个网管系统上传的故障预测模型，其中，每个网管系统基于网管系统对应的基站样本数据进行模型训练得到网管系统对应的故障预测模型；然后，在多个故障预测模型中确定至少一个候选预测模型，并基于至少一个候选预测模型确定目标预测模型；最后，将目标预测模型下发至多个网管系统，以使多个网管系统利用目标预测模型对对应的基站进行故障预测。从而实现对基站设备故障的预测，提前发现故障问题，减少损失，为用户提供更加稳定的通信服务。并且，本公开中通过聚合不同网管系统的本地模型(即候选预测模型)，形成最终的联合模型(即目标预测模型)，并利用联合模型预测基站故障，可以避免由于本地数据的局限性影响故障预测的准确性。
在本公开又一实施例中，以故障预测模型为BBU设备过温预测模型为例，网管系统可以通过以下方式训练故障预测模型。
(1)对基站样本数据进行归一化处理，即将一个BBU需要的数据处理成一行，形成最终的样本特征数据如下
(2)根据实际BBU温度环境健康度的情况对样本特征数据打标签
(3)基于样本特征数据和特征函数(1)，采用梯度下降的方式进行训练，其中，特征函数(1)如下：
其中，x1...xn代表步骤(1)中a到i所表示的特征变量，y为预期结果，c1...cn代表a到i所表示的特征变量的特征参数，c0为常量参数。
给定初始步长和初始模型，然后利用梯度下降的方式不断更新模型，当损失函数随着梯度下降后波动非常小，代表其已经基本收敛至最小值，模型训练完毕，此时c0，c1，c2...cn即为训练出来的故障预测模型的参数。
在本公开又一实施例中，如图3所示，S102，可以包括以下步骤S201至步骤S203。
S201，针对每个候选预测模型，获取候选预测模型对应的权重，以及，获取候选预测模型的模型参数，其中，至少一个候选预测模型为基于同一个预设算法训练得到。
S202，基于至少一个候选预测模型对应的模型参数和权重进行加权平均运算，得到目标参数。
S203，利用目标参数对预设算法进行配置，得到目标预测模型。
在本公开实施例中，候选预测模型为基于同一个预设算法(如特征函数(1))训练得到。可以设置每个候选预测模型的权重，利用该权重对所有候选预测模型对应的模型参数进行加权平均运算，得到目标参数，其中，当模型参数有多个时，则针对每个模型参数进行加权平均运算，得到对应的目标参数。进而利用目标参数对预设算法进行配置，得到目标预测模型。
例如，预设算法中包括参数a和b，候选预测模型有模型一和模型二，其中模型一的参数a的值为1，参数b的值为2，其权重为0.3；模型二的参数a的值为3，参数b的值为4，其权重为0.7；则目标预测模型的参数a的值为：1*0.3+3*0.7＝2.4，目标预测模型的参数b的值为：2*0.3+4*0.7＝3.4。
作为一些实施方式，获取候选预测模型对应的权重，可以包括以下步骤：获取候选预测模型对应的故障样本数量，以及，至少一个候选预测模型对应的故障样本总量；计算故障样本数量与故障样本总量的比值，将比值作为候选预测模型对应的权重。
每个网管系统在上传故障预测模型时，还可以同时上传训练该故障预测模型的故障样本数量(即风险和故障态的样本的数量)，对所有候选预测模型的故障样本数量求和，得到所有候选预测模型的故障样本总量。进而，针对每个候选预测模型，将其对应的故障样本数量与故障样本总量的比值，作为该候选预测模型对应的权重。
例如，网管系统T1，T2，T3上传故障样本数量分别为M1，M2，M3，则网管系统T1 对应的权重为V1＝M1/(M1+M2+M3)，网管系统T2对应的权重为V2＝M2/(M1+M2+M3)，网管系统T3对应的权重为V3＝M3/(M1+M2+M3)。
通过本方案，可以利用候选预测模型对应的权重对模型参数进行加权平均运算，得到目标参数，进而利用目标参数配置预设算法得到目标预测模型。从而提高目标预测模型的预测准确性。
在本公开又一实施例中，可以预设每个候选预测模型的权重相同，则直接计算模型参数的平均值即可得到目标参数，从而减少计算难度。
在本公开又一实施例中，该方法还可以包括以下步骤：接收多个网管系统发送的预测结果，其中，每个网管系统利用目标预测模型对对应的基站进行故障预测得到基站对应的预测结果；以及基于多个预测结果，生成基站故障报告。
在本公开实施例中，每个网管系统可以将实时采集的特征数据输入目标预测模型中，得到对应的预测结果，预测结果一般包括三种情况：正常、风险及故障。进而，将预测结果上传至服务器，服务器可以根据多个网管系统上传的预测结果，生成基站故障报告，其中，基站故障报告中包括每个基站的基站标识和预测结果。可选的，基站故障报告中还可以包括对应的故障问题和处理方式。从而方便工作人员查看多个基站的健康状况。
基于相同的技术构思，本公开实施例还提供了一种故障预测装置，如图4所示，该装置包括：获取模块301，被配置为获取多个网管系统上传的故障预测模型，其中，每个网管系统基于网管系统对应的基站样本数据进行模型训练得到网管系统对应的故障预测模型；确定模块302，被配置为在多个故障预测模型中确定至少一个候选预测模型，并基于至少一个候选预测模型确定目标预测模型；以及下发模块303，被配置为将目标预测模型下发至多个网管系统，以使多个网管系统利用目标预测模型对对应的基站进行故障预测。
在一些可能的实施方式中，确定模块，被配置为：针对每个候选预测模型，获取候选预测模型对应的权重，以及，获取候选预测模型的模型参数，其中，至少一个候选预测模型为基于同一个预设算法训练得到；基于至少一个候选预测模型对应的模型参数和权重进行加权平均运算，得到目标参数；以及利用目标参数对预设算法进行配置，得到目标预测模型。
在一些可能的实施方式中，确定模块，还被配置为：获取候选预测模型对应的故障样本数量，以及，至少一个候选预测模型对应的故障样本总量；以及计算故障样本数量与故障样本总量的比值，将比值作为候选预测模型对应的权重。
在一些可能的实施方式中，确定模块，还被配置为：针对多个故障预测模型中的任一故障预测模型，获取故障预测模型对应的样本数量；以及将样本数量大于或等于预设数量阈值的故障预测模型，确定为候选预测模型。
在一些可能的实施方式中，获取模块，被配置为：针对多个网管系统中的任一网管系统，获取网管系统上传的第一加密数据，以及，确定网管系统对应的第一加密规则，其中，网管系统基于第一加密规则将对应的故障预测模型进行加密得到第一加密数据；以及基于第一加密规则对第一加密数据进行解密，得到网管系统对应的故障预测模型。
在一些可能的实施方式中，下发模块，被配置为：基于第二加密规则对目标预测模型进行加密，得到第二加密数据；以及将第二加密数据下发至多个网管系统，以使多个网管系统基于第二加密规则对第二加密数据进行解密得到目标预测模型。
在一些可能的实施方式中，装置还包括生成模块，被配置为：接收多个网管系统发送的预测结果，其中，每个网管系统利用目标预测模型对对应的基站进行故障预测得到基站对应的预测结果；以及基于多个预测结果，生成基站故障报告。
本公开实施例中，首先，获取多个网管系统上传的故障预测模型，其中，每个网管系统基于网管系统对应的基站样本数据进行模型训练得到网管系统对应的故障预测模型；然后，在多个故障预测模型中确定至少一个候选预测模型，并基于至少一个候选预测模型确定目标预测模型；最后，将目标预测模型下发至多个网管系统，以使多个网管系统利用目标预测模型对对应的基站进行故障预测。从而实现对基站设备故障的预测，提前发现故障问题，减少损失，为用户提供更加稳定的通信服务。并且，本公开中通过聚合不同网管系统的本地模型(即候选预测模型)，形成最终的联合模型(即目标预测模型)，并利用联合模型预测基站故障，可以避免由于本地数据的局限性影响故障预测的准确性。
基于相同的技术构思，本公开实施例还提供了一种电子设备，如图5所示，包括处理器111、通信接口112、存储器113和通信总线114。处理器111，通信接口112，存储器113通过通信总线114完成相互间的通信。存储器113，用于存放计算机程序。处理器111，用于执行存储器113上所存放的程序时，实现如下步骤：获取多个网管系统上传的故障预测模型，其中，每个网管系统基于网管系统对应的基站样本数据进行模型训练得到网管系统对应的故障预测模型；在多个故障预测模型中确定至少一个候选预测模型，并基于至少一个候选预测模型确定目标预测模型；以及将目标预测模型下发至多个网管系统，以使多个网管系统利用目标预测模型对对应的基站进行故障预测。
上述电子设备提到的通信总线可以是外设部件互连标准(Peripheral Component Interconnect，PCI)总线或扩展工业标准结构(Extended Industry Standard Architecture，EISA)总线等。该通信总线可以分为地址总线、数据总线、控制总线等。为便于表示，图中仅用一条粗线表示，但并不表示仅有一根总线或一种类型的总线。
通信接口用于上述电子设备与其他设备之间的通信。
存储器可以包括随机存取存储器(Random Access Memory，RAM)，也可以包括非易 失性存储器(Non-Volatile Memory，NVM)，例如至少一个磁盘存储器。可选的，存储器还可以是至少一个位于远离前述处理器的存储装置。
上述的处理器可以是通用处理器，包括中央处理器(Central Processing Unit，CPU)、网络处理器(Network Processor，NP)等；还可以是数字信号处理器(Digital Signal Processing，DSP)、专用集成电路(Application Specific Integrated Circuit，ASIC)、现场可编程门阵列(Field-Programmable Gate Array，FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件。
在本公开提供的又一实施例中，还提供了一种计算机可读存储介质，该计算机可读存储介质内存储有计算机程序，计算机程序被处理器执行时实现上述任一故障预测方法的步骤。
在本公开提供的又一实施例中，还提供了一种包含指令的计算机程序产品，当其在计算机上运行时，使得计算机执行上述实施例中任一故障预测方法。
在上述实施例中，可以全部或部分地通过软件、硬件、固件或者其任意组合来实现。当使用软件实现时，可以全部或部分地以计算机程序产品的形式实现。计算机程序产品包括一个或多个计算机指令。在计算机上加载和执行计算机程序指令时，全部或部分地产生按照本公开实施例的流程或功能。计算机可以是通用计算机、专用计算机、计算机网络、或者其他可编程装置。计算机指令可以存储在计算机可读存储介质中，或者从一个计算机可读存储介质向另一个计算机可读存储介质传输，例如，计算机指令可以从一个网站站点、计算机、服务器或数据中心通过有线(例如同轴电缆、光纤、数字用户线(DSL))或无线(例如红外、无线、微波等)方式向另一个网站站点、计算机、服务器或数据中心进行传输。计算机可读存储介质可以是计算机能够存取的任何可用介质或者是包含一个或多个可用介质集成的服务器、数据中心等数据存储设备。可用介质可以是磁性介质，(例如，软盘、硬盘、磁带)、光介质(例如，DVD)、或者半导体介质(例如固态硬盘Solid State Disk(SSD))等。
本公开实施例提供了一种故障预测方法、装置、电子设备及存储介质，本公开通过，首先，获取多个网管系统上传的故障预测模型，其中，每个网管系统基于网管系统对应的基站样本数据进行模型训练得到网管系统对应的故障预测模型；然后，在多个故障预测模型中确定至少一个候选预测模型，并基于至少一个候选预测模型确定目标预测模型；最后，将目标预测模型下发至多个网管系统，以使多个网管系统利用目标预测模型对对应的基站进行故障预测。从而实现对基站设备故障的预测，提前发现故障问题，减少损失，为用户提供更加稳定的通信服务。并且，本公开中通过聚合不同网管系统的本地模型(即候选预测模型)，形成最终的联合模型(即目标预测模型)，并利用联合模型预测基站故障，可以 避免由于本地数据的局限性影响故障预测的准确性。当然，实施本公开的任一产品或方法并不一定需要同时达到以上的所有优点。
需要说明的是，在本文中，诸如“第一”和“第二”等之类的关系术语仅仅用来将一个实体或者操作与另一个实体或操作区分开来，而不一定要求或者暗示这些实体或操作之间存在任何这种实际的关系或者顺序。而且，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、物品或者设备不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、物品或者设备所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括所述要素的过程、方法、物品或者设备中还存在另外的相同要素。
以上所述仅是本公开的具体实施方式，使本领域技术人员能够理解或实现本公开。对这些实施例的多种修改对本领域的技术人员来说将是显而易见的，本文中所定义的一般原理可以在不脱离本公开的精神或范围的情况下，在其它实施例中实现。因此，本公开将不会被限制于本文所示的这些实施例，而是要符合与本文所公开的原理和新颖特点相一致的最宽的范围。

Claims (10)

  1. 一种故障预测方法，包括：
    获取多个网管系统上传的故障预测模型，其中，每个网管系统基于所述网管系统对应的基站样本数据进行模型训练得到所述网管系统对应的故障预测模型；
    在多个所述故障预测模型中确定至少一个候选预测模型，并基于至少一个所述候选预测模型确定目标预测模型；以及
    将所述目标预测模型下发至多个所述网管系统，以使多个所述网管系统利用所述目标预测模型对对应的基站进行故障预测。
  2. 根据权利要求1所述的方法，其中，所述基于至少一个所述候选预测模型确定目标预测模型，包括：
    针对每个候选预测模型，获取所述候选预测模型对应的权重，以及，获取所述候选预测模型的模型参数，其中，至少一个所述候选预测模型为基于同一个预设算法训练得到；
    基于至少一个所述候选预测模型对应的模型参数和权重进行加权平均运算，得到目标参数；以及
    利用所述目标参数对所述预设算法进行配置，得到所述目标预测模型。
  3. 根据权利要求2所述的方法，其中，所述获取所述候选预测模型对应的权重，包括：
    获取所述候选预测模型对应的故障样本数量，以及，至少一个所述候选预测模型对应的故障样本总量；以及
    计算所述故障样本数量与所述故障样本总量的比值，将所述比值作为所述候选预测模型对应的权重。
  4. 根据权利要求1所述的方法，其中，所述在多个所述故障预测模型中确定至少一个候选预测模型，包括：
    针对多个所述故障预测模型中的任一故障预测模型，获取所述故障预测模型对应的样本数量；以及
    将所述样本数量大于或等于预设数量阈值的故障预测模型，确定为所述候选预测模型。
  5. 根据权利要求1所述的方法，其中，所述获取多个网管系统上传的故障预测模型，包括：
    针对多个所述网管系统中的任一网管系统，获取所述网管系统上传的第一加密数据，以及，确定所述网管系统对应的第一加密规则，其中，所述网管系统基于所述第一加密规则将对应的故障预测模型进行加密得到所述第一加密数据；以及
    基于所述第一加密规则对所述第一加密数据进行解密，得到所述网管系统对应的故障预测模型。
  6. 根据权利要求1所述的方法，其中，所述将所述目标预测模型下发至多个所述网管系统，包括：
    基于第二加密规则对所述目标预测模型进行加密，得到第二加密数据；以及
    将所述第二加密数据下发至多个所述网管系统，以使多个所述网管系统基于所述第二加密规则对所述第二加密数据进行解密得到所述目标预测模型。
  7. 根据权利要求1所述的方法，还包括：
    接收多个所述网管系统发送的预测结果，其中，每个所述网管系统利用所述目标预测模型对对应的基站进行故障预测得到所述基站对应的预测结果；以及
    基于多个所述预测结果，生成基站故障报告。
  8. 一种故障预测装置，包括：
    获取模块，被配置为获取多个网管系统上传的故障预测模型，其中，每个网管系统基于所述网管系统对应的基站样本数据进行模型训练得到所述网管系统对应的故障预测模型；
    确定模块，被配置为在多个所述故障预测模型中确定至少一个候选预测模型，并基于至少一个所述候选预测模型确定目标预测模型；以及
    下发模块，被配置为将所述目标预测模型下发至多个所述网管系统，以使多个所述网管系统利用所述目标预测模型对对应的基站进行故障预测。
  9. 一种电子设备，包括处理器、通信接口、存储器和通信总线，其中，处理器、通信接口、存储器通过通信总线完成相互间的通信；
    存储器，用于存储计算机程序；以及
    处理器，用于执行存储器上所存储的程序时，实现权利要求1-7中任一所述的方法。
  10. 一种计算机可读存储介质，其中，所述计算机可读存储介质内存储有计算机程序，所述计算机程序被处理器执行时实现权利要求1-7中任一所述的方法。
PCT/CN2023/072164 2022-06-13 2023-01-13 故障预测方法、装置、电子设备及存储介质 WO2023241042A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210666827.0 2022-06-13
CN202210666827.0A CN117279014A (zh) 2022-06-13 2022-06-13 故障预测方法、装置、电子设备及存储介质

Publications (1)

Publication Number Publication Date
WO2023241042A1 true WO2023241042A1 (zh) 2023-12-21

Family

ID=89193084

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/072164 WO2023241042A1 (zh) 2022-06-13 2023-01-13 故障预测方法、装置、电子设备及存储介质

Country Status (2)

Country Link
CN (1) CN117279014A (zh)
WO (1) WO2023241042A1 (zh)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109871702A (zh) * 2019-02-18 2019-06-11 深圳前海微众银行股份有限公司 联邦模型训练方法、系统、设备及计算机可读存储介质
CN111144950A (zh) * 2019-12-30 2020-05-12 北京顺丰同城科技有限公司 模型筛选方法、装置、电子设备及存储介质
CN114282691A (zh) * 2021-10-27 2022-04-05 腾讯科技（深圳）有限公司 一种联邦学习方法、装置、设备、存储介质及计算机程序
CN114648050A (zh) * 2020-12-02 2022-06-21 新智数字科技有限公司 设备故障预测方法、装置、可读存储介质及电子设备

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109871702A (zh) * 2019-02-18 2019-06-11 深圳前海微众银行股份有限公司 联邦模型训练方法、系统、设备及计算机可读存储介质
CN111144950A (zh) * 2019-12-30 2020-05-12 北京顺丰同城科技有限公司 模型筛选方法、装置、电子设备及存储介质
CN114648050A (zh) * 2020-12-02 2022-06-21 新智数字科技有限公司 设备故障预测方法、装置、可读存储介质及电子设备
CN114282691A (zh) * 2021-10-27 2022-04-05 腾讯科技（深圳）有限公司 一种联邦学习方法、装置、设备、存储介质及计算机程序

Also Published As

Publication number Publication date
CN117279014A (zh) 2023-12-22

Similar Documents

Publication Publication Date Title
WO2021244211A1 (zh) 区块链消息处理方法、装置、计算机以及可读存储介质
US9998340B2 (en) Method and system to monitor a network
US10581697B2 (en) SDN controlled PoE management system
WO2016000464A1 (zh) 一种支持集中化北向接口对接的方法及系统
CN114629729B (zh) 一种多表集抄系统、方法、电子设备及存储介质
CN111654321A (zh) 卫星网络的管理方法、装置及电子设备
WO2019137052A1 (zh) 网络运维的方法及装置
WO2023241042A1 (zh) 故障预测方法、装置、电子设备及存储介质
JP5575271B2 (ja) 通信システム内でリソース使用を制御する方法
CN108259207A (zh) 流量统计方法、客户端、服务器及系统
WO2020056633A1 (zh) 一种预测网络速率的方法及预测装置
US20200076927A1 (en) Virtualization with distributed adaptive message brokering
WO2023082871A1 (zh) 流业务建模方法、装置、平台、电子设备和存储介质
CN114243914B (zh) 电力监控系统
TWI718778B (zh) 聯網裝置及其流量預估方法
US8868709B2 (en) Programmatically selecting a service provider based on assured quality of service attributes
EP3205138A1 (en) Method, apparatus and system for changing a network based on received network information
CN109756384B (zh) 数据源状态判定方法及装置、计算机可读存储介质
CN111163088B (zh) 消息处理方法、系统、装置及电子设备
CN109960924A (zh) 一种子系统登录方法、装置、系统及电子设备
WO2024051805A1 (zh) 一种运维网络自治等级的评估方法、装置及系统
Lin et al. A multi-state computer network within transmission error rate and time constraints
CN113595240B (zh) 电力数据的检测方法、装置、设备及存储介质
TWI700939B (zh) 網路設備的品質監控伺服器及其方法
US20230353648A1 (en) Data tracking for data owners

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23822614

Country of ref document: EP

Kind code of ref document: A1