CN118134513A - Network element equipment depreciation prediction method, device and equipment - Google Patents

Abstract

The application discloses a network element equipment depreciation prediction method, a device, electronic equipment and a storage medium, comprising the following steps: acquiring historical performance data and corresponding historical alarm numbers of a target network element; training a preset long-short-period memory network model based on the historical performance data and the historical alarm number, and iteratively adjusting model parameters of the preset long-short-period memory network model to obtain a network element service life prediction model; after the current performance data and the current alarm number of the target network element are obtained, the current performance data and the current alarm number are input into a network element life prediction model, and the residual available time length of the target network element from the current time is predicted to obtain the predicted life of target network element equipment; and calculating the current depreciation cost of the target network element according to the predicted service life. Therefore, the residual service life of the target network element can be predicted more accurately, and the obtained current depreciation cost can be more in line with the actual use condition of the target network element, so that the accuracy of the cost estimation of the target network element is further.

Description

Network element equipment depreciation prediction method, device and equipment

Technical Field

The application belongs to the field of data processing, and particularly relates to a network element equipment depreciation prediction method, device, equipment and storage medium.

Background

The 5G base station is a key facility supporting the 5G network for providing high-speed, low-delay data transmission and connection, and with the rapid development of 5G communication technology, the construction and operation of the 5G base station are becoming more and more important.

Due to the complexity of the 5G technology and the high integration of the 5G base station, the construction cost of the 5G base station is significantly increased, and therefore, the importance of depreciation cost is significantly increased in the benefit evaluation of the 5G base station. The depreciation cost means that the value of the 5G base station gradually decreases along with the time, reflects the consumption of the service life of the base station and the requirement of technical upgrading, and can help operators and related stakeholders to determine the economic life, the return on investment rate and the time for updating and maintaining equipment of the base station.

Therefore, it is necessary to accurately calculate the resource depreciation cost of the 5G base station to accurately evaluate the benefit of the 5G base station and make a decision according to the situation. However, the existing method for calculating the depreciation cost of the base station mainly adopts the traditional method for calculating the depreciation cost of the fixed asset, and does not consider the difference and the change between different base stations, so that the calculation result of the depreciation cost is not accurate enough.

Disclosure of Invention

The embodiment of the application aims to provide a network element equipment depreciation prediction method, a device, equipment and a storage medium, which can solve the problem that the calculation result of depreciation cost is not accurate because the traditional fixed asset depreciation method is adopted at present and the difference and the change among different base stations are not considered.

In a first aspect, an embodiment of the present application provides a method for predicting depreciation of network element equipment, including:

acquiring historical performance data and corresponding historical alarm numbers of a target network element;

Training a preset long-short-period memory network model based on the historical performance data and the historical alarm number, and iteratively adjusting model parameters of the preset long-short-period memory network model to obtain a network element life prediction model;

After the current performance data and the current alarm number of the target network element are obtained, the current performance data and the current alarm number are input into the network element life prediction model, and the residual available time length of the target network element from the current time is predicted to obtain the predicted life of the target network element equipment;

and calculating the current depreciation cost of the target network element according to the predicted service life.

Optionally, training the preset long-short-period memory network model based on the historical performance data and the historical alarm number, and iteratively adjusting model parameters of the preset long-short-period memory network model to obtain a network element life prediction model, including:

Acquiring type information of the target network element, and determining an alarm number threshold corresponding to the target network element based on the type information;

Marking the historical alarm number and the corresponding historical performance data as negative samples when the historical alarm number is larger than or equal to the alarm number threshold value, and marking the historical alarm number and the corresponding historical performance data as positive samples when the historical alarm number is smaller than the alarm number threshold value, so as to obtain training data;

and training a preset long-short-period memory network model based on the training data, and iteratively adjusting model parameters of the preset long-short-period memory network model to obtain a network element life prediction model.

Optionally, training the preset long-short-term memory network model based on the training data, iteratively adjusting model parameters of the preset long-short-term memory network model to obtain a network element life prediction model, including:

inputting the training data into a preset long-term and short-term memory network model, and predicting the remaining available time length of the target network element from the corresponding time of the training data to obtain the service life of the target network element to be verified;

Comparing the service life to be verified with the labeling result of the training data, and determining the loss value of the preset long-period and short-period memory network model;

And under the condition that the loss value is larger than a preset loss threshold value, iteratively adjusting the model parameters of the preset long-short-period memory network model until the loss value is smaller than or equal to the preset loss threshold value, wherein the adjusted preset long-short-period memory network model is used as a network element life prediction model.

Optionally, inputting the training data into a preset long-short-term memory network model, predicting the remaining available time length of the target network element from the time corresponding to the training data, to obtain the service life to be verified of the target network element, including:

And carrying out normalization processing and normalization processing on the training data, inputting the processed training data into a preset long-term and short-term memory network model, and predicting the residual available time length of the target network element from the corresponding time of the training data to obtain the service life to be verified of the target network element.

Optionally, the determining the loss value of the preset long-term and short-term memory network model by comparing the life to be verified with the labeling result of the training data includes:

And calculating an average square error between the service life to be verified and the labeling result of the training data, and taking the average square error as a loss value of the preset long-short-term memory network model.

Optionally, the inputting the current performance data and the current alarm number into the network element lifetime prediction model predicts a remaining available time length of the target network element from a current time to obtain a predicted lifetime of the target network element device, including:

Inputting the current performance data and the current alarm number into the network element life prediction model, and splicing the current performance data and the current alarm number by an input layer to obtain input layer data;

Modeling the time dependency relationship in the input layer data by using a long-term and short-term memory network layer structure to obtain characteristic data;

and the full connection layer converts and overfits the feature space dimension of the feature data to obtain the predicted service life of the target network element equipment.

Optionally, the full connection layer performs conversion and overfitting processing on the feature space dimension of the feature data to obtain the predicted lifetime of the target network element device, including:

the full connection layer converts and overfits the feature space dimension of the feature data to obtain the predicted alarm number of the target network element equipment from the current time;

comparing the predicted alarm number with an alarm number threshold corresponding to the target network element, and determining an unavailable time point of the target network element equipment;

And calculating the difference value between the unavailable time point and the current time to serve as the predicted service life of the target network element equipment.

Optionally, the calculating the current depreciation cost of the target network element according to the predicted lifetime includes:

Acquiring a fixed asset original value of the target network element;

and taking the ratio of the original value of the fixed asset to the predicted life as the current depreciation cost of the target network element.

Optionally, the step of using the ratio of the fixed asset primary value to the predicted lifetime as the current depreciation cost of the target network element includes:

Taking the ratio of the fixed asset original value to the predicted life as the candidate current depreciation cost of the target network element;

Acquiring the historical depreciation cost of the target network element;

And calculating a difference value between the original value of the fixed asset and the historical depreciation cost, and taking the minimum value of the difference value and the candidate current depreciation cost as the current depreciation cost of the target network element.

In a second aspect, an embodiment of the present application provides a network element device depreciation prediction apparatus, including:

the acquisition module is used for acquiring the historical performance data of the target network element and the corresponding historical alarm number;

The training module is used for training a preset long-short-period memory network model based on the historical performance data and the historical alarm number, and iteratively adjusting model parameters of the preset long-short-period memory network model to obtain a network element service life prediction model;

The prediction module is used for inputting the current performance data and the current alarm number into the network element life prediction model after acquiring the current performance data and the current alarm number of the target network element, and predicting the residual available time length of the target network element from the current time to obtain the predicted life of the target network element equipment;

And the calculating module is used for calculating the current depreciation cost of the target network element according to the predicted service life.

Optionally, the training module is configured to:

Acquiring type information of the target network element, and determining an alarm number threshold corresponding to the target network element based on the type information;

Marking the historical alarm number and the corresponding historical performance data as negative samples when the historical alarm number is larger than or equal to the alarm number threshold value, and marking the historical alarm number and the corresponding historical performance data as positive samples when the historical alarm number is smaller than the alarm number threshold value, so as to obtain training data;

and training a preset long-short-period memory network model based on the training data, and iteratively adjusting model parameters of the preset long-short-period memory network model to obtain a network element life prediction model.

Optionally, the training module is configured to:

inputting the training data into a preset long-term and short-term memory network model, and predicting the remaining available time length of the target network element from the corresponding time of the training data to obtain the service life of the target network element to be verified;

Comparing the service life to be verified with the labeling result of the training data, and determining the loss value of the preset long-period and short-period memory network model;

And under the condition that the loss value is larger than a preset loss threshold value, iteratively adjusting the model parameters of the preset long-short-period memory network model until the loss value is smaller than or equal to the preset loss threshold value, wherein the adjusted preset long-short-period memory network model is used as a network element life prediction model.

Optionally, the training module is configured to:

And carrying out normalization processing and normalization processing on the training data, inputting the processed training data into a preset long-term and short-term memory network model, and predicting the residual available time length of the target network element from the corresponding time of the training data to obtain the service life to be verified of the target network element.

Optionally, the training module is configured to:

And calculating an average square error between the service life to be verified and the labeling result of the training data, and taking the average square error as a loss value of the preset long-short-term memory network model.

Optionally, the prediction module is configured to:

Inputting the current performance data and the current alarm number into the network element life prediction model, and splicing the current performance data and the current alarm number by an input layer to obtain input layer data;

Modeling the time dependency relationship in the input layer data by using a long-term and short-term memory network layer structure to obtain characteristic data;

and the full connection layer converts and overfits the feature space dimension of the feature data to obtain the predicted service life of the target network element equipment.

Optionally, the prediction module is configured to:

the full connection layer converts and overfits the feature space dimension of the feature data to obtain the predicted alarm number of the target network element equipment from the current time;

comparing the predicted alarm number with an alarm number threshold corresponding to the target network element, and determining an unavailable time point of the target network element equipment;

And calculating the difference value between the unavailable time point and the current time to serve as the predicted service life of the target network element equipment.

Optionally, the computing module is configured to:

Acquiring a fixed asset original value of the target network element;

and taking the ratio of the original value of the fixed asset to the predicted life as the current depreciation cost of the target network element.

Optionally, the computing module is configured to:

Taking the ratio of the fixed asset original value to the predicted life as the candidate current depreciation cost of the target network element;

Acquiring the historical depreciation cost of the target network element;

And calculating a difference value between the original value of the fixed asset and the historical depreciation cost, and taking the minimum value of the difference value and the candidate current depreciation cost as the current depreciation cost of the target network element.

In a third aspect, an embodiment of the present application provides an electronic device comprising a processor and a memory storing a program or instructions executable on the processor, which when executed by the processor, implement the steps of the method as described in the first aspect.

In a fourth aspect, embodiments of the present application provide a readable storage medium having stored thereon a program or instructions which when executed by a processor perform the steps of the method according to the first aspect.

In a fifth aspect, an embodiment of the present application provides a chip, where the chip includes a processor and a communication interface, where the communication interface is coupled to the processor, and where the processor is configured to execute a program or instructions to implement a method according to the first aspect.

In a sixth aspect, embodiments of the present application provide a computer program product stored in a storage medium, the program product being executable by at least one processor to implement the method according to the first aspect.

In the application, the historical performance data and the corresponding historical alarm number of the target network element are obtained; training a preset long-short-period memory network model based on the historical performance data and the historical alarm number, and iteratively adjusting model parameters of the preset long-short-period memory network model to obtain a network element service life prediction model; after the current performance data and the current alarm number of the target network element are obtained, the current performance data and the current alarm number are input into a network element life prediction model, and the residual available time length of the target network element from the current time is predicted to obtain the predicted life of target network element equipment; and calculating the current depreciation cost of the target network element according to the predicted service life.

In this way, the service life of the target network element is predicted based on the current alarm number of the current performance data set through the trained network element service life prediction model, and further, the corresponding depreciation cost is calculated based on the residual service life of the target network element.

Drawings

Fig. 1 is a flow chart illustrating a method of network element device depreciation prediction according to an exemplary embodiment;

FIG. 2 is a diagram illustrating a training process for a network element lifetime prediction model, according to an example embodiment;

fig. 3 is a schematic step diagram illustrating a specific implementation of a network element device depreciation prediction method according to an exemplary embodiment;

FIG. 4 is a schematic diagram of a model structure of a network element lifetime prediction model, according to an example embodiment;

Fig. 5 is a block diagram illustrating a network element device depreciation prediction apparatus according to an exemplary embodiment;

FIG. 6 is a block diagram of a base station terminal accessing an electronic device, according to an example embodiment;

Fig. 7 is a block diagram illustrating an apparatus for terminal access according to an example embodiment.

Detailed Description

The technical solutions of the embodiments of the present application will be clearly described below with reference to the drawings in the embodiments of the present application, and it is apparent that the described embodiments are some embodiments of the present application, but not all embodiments. All other embodiments, which are obtained by a person skilled in the art based on the embodiments of the present application, fall within the scope of protection of the present application.

The terms first, second and the like in the description and in the claims, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged, as appropriate, such that embodiments of the present application may be implemented in sequences other than those illustrated or described herein, and that the objects identified by "first," "second," etc. are generally of a type, and are not limited to the number of objects, such as the first object may be one or more. Furthermore, in the description and claims, "and/or" means at least one of the connected objects, and the character "/", generally means that the associated object is an "or" relationship.

The network element equipment depreciation prediction method provided by the embodiment of the application is described in detail below through specific embodiments and application scenarios thereof with reference to the accompanying drawings.

Fig. 1 is a flowchart illustrating a network element device depreciation prediction method according to an exemplary embodiment, the network element device depreciation prediction method including the following steps.

In step S11, historical performance data of the target network element and a corresponding historical alert number are obtained.

The 5G base station is a key facility supporting the 5G network for providing high-speed, low-delay data transmission and connection, and with the rapid development of 5G communication technology, the construction and operation of the 5G base station are becoming more and more important.

Due to the complexity of the 5G technology and the high integration of the 5G base station, the construction cost of the 5G base station is significantly increased, and therefore, the importance of depreciation cost is significantly increased in the benefit evaluation of the 5G base station. The depreciation cost means that the value of the 5G base station gradually decreases along with the time, reflects the consumption of the service life of the base station and the requirement of technical upgrading, and can help operators and related stakeholders to determine the economic life, the return on investment rate and the time for updating and maintaining equipment of the base station.

Therefore, it is necessary to accurately calculate the resource depreciation cost of the 5G base station to accurately evaluate the benefit of the 5G base station and make a decision according to the situation. However, the existing method for calculating the depreciation cost of the base station mainly adopts the traditional method for calculating the depreciation cost of the fixed asset, and does not consider the difference and the change between different base stations, so that the calculation result of the depreciation cost is not accurate enough.

Therefore, there is a strong need for an efficient, highly reliable and refined network element equipment depreciation prediction method, based on which the present invention provides a lifetime prediction algorithm based on LSTM (Long Short-Term Memory network), which can calculate depreciation cost of the network element equipment more accurately, so as to solve the above problems pointedly.

The network element device includes a base station in the 5G network, and may also include other devices such as a network controller and a network service node, which is not limited in this embodiment of the present application.

In this step, firstly, the historical performance data and the corresponding historical alarm number of the target network element are obtained, wherein the performance data can embody the equipment load condition of the target network element and is a high-order characteristic affecting the degradation level of the network element.

In one implementation, the historical performance data of the target network element includes, but is not limited to:

PRB (Probability of Reception of a Block, block reception probability) utilization, downlink PDCP (PACKET DATA Convergence Protocol ) traffic, uplink PDCP traffic, RRC (Radio Resource Control ) connection user number.

In the application, the alarm number can be used as a key index of whether the target network element is available, and it can be understood that when the target network element is available, the equipment operates normally, and the error reporting is less, so that the alarm number is less, otherwise, when the target network element is unavailable, the equipment operates abnormally gradually, so that more alarm numbers are generated.

In step S12, training the preset long-short-period memory network model based on the historical performance data and the historical alarm number, and iteratively adjusting model parameters of the preset long-short-period memory network model to obtain a network element life prediction model.

In the step, a preset LSTM model is selected as a network element life prediction model for training, and the LSTM model is used as a supervised learning algorithm model based on a time sequence, so that data characteristics of a target network element during working can be more fully learned, a finer modeling effect than that of manual modeling can be obtained, and the accuracy of a network element life prediction result can be improved.

It will be appreciated that other methods commonly used in the art of life prediction include: machine learning methods represented by neural networks and deep learning, and traditional statistical data driving methods represented by Wiener processes and Gamma processes.

The latter can estimate the degradation model parameters according to the degradation track and deduce the analysis probability distribution of the residual life, but the residual life prediction accuracy is greatly influenced by the selected degradation model. The LSTM model selected by the application is used as a machine learning method, can extract effective information contained in the monitoring data, and is used for describing the nonlinear relation between the characteristic information and the residual life, so that the method has certain universality in the field of residual life prediction.

In one implementation, training a preset long-short-period memory network model based on historical performance data and historical alarm numbers, iteratively adjusting model parameters of the preset long-short-period memory network model to obtain a network element life prediction model, including:

Acquiring type information of a target network element, and determining an alarm number threshold corresponding to the target network element based on the type information;

Marking the historical alarm number and corresponding historical performance data as negative samples when the historical alarm number is larger than or equal to an alarm number threshold value, and marking the historical alarm number and corresponding historical performance data as positive samples when the historical alarm number is smaller than the alarm number threshold value to obtain training data;

training the preset long-short-period memory network model based on the training data, and iteratively adjusting model parameters of the preset long-short-period memory network model to obtain the network element life prediction model.

It can be understood that the device type corresponding to the target network element is also a high-order feature affecting the degradation level of the network element, so that the device of the target network element can be classified into two types, i.e. indoor and outdoor, according to the device installation type, and meanwhile, can be classified into several different types, i.e. AAU (ACTIVE ANTENNA Unit ), RRU (Remote Radio Unit, remote radio Unit) and the like, according to the network element type.

Further, by taking the number of alarms when the target network element is unavailable as a standard, taking the type of the network element and the installation type as dimensions, calculating the average unavailable number of alarms of different types of network elements as corresponding alarm number thresholds, and then carrying out data annotation on training data by using the calculated alarm number thresholds, thereby being used for training a preset LSTM model.

In one implementation, training the preset long-short-period memory network model based on training data, iteratively adjusting model parameters of the preset long-short-period memory network model to obtain a network element life prediction model, including:

Inputting training data into a preset long-term memory network model, and predicting the remaining available time length of the target network element from the corresponding time of the training data to obtain the service life of the target network element to be verified; comparing the service life to be verified with the labeling result of the training data, and determining a loss value of a preset long-period memory network model; and under the condition that the loss value is larger than a preset loss threshold value, iteratively adjusting model parameters of the preset long-short-period memory network model until the loss value is smaller than or equal to the preset loss threshold value, wherein the adjusted preset long-short-period memory network model is used as a network element life prediction model.

That is, training data is input into a preset LSTM model to train, the preset LSTM model is converged based on a preset loss function and an optimization algorithm, and after the preset LSTM model is converged, the trained model is durable and used as a network element life prediction model for a subsequent target network element life prediction flow.

In one implementation, the method for inputting training data into a preset long-term and short-term memory network model, predicting the remaining available time of a target network element from the corresponding time of the training data to obtain the service life of the target network element to be verified includes:

and carrying out normalization processing and normalization processing on the training data, inputting the processed training data into a preset long-short-term memory network model, and predicting the remaining available time length of the target network element from the corresponding time of the training data to obtain the service life of the target network element to be verified.

That is, the training data needs to be further preprocessed before being input into the preset long-term memory network model, wherein the preprocessing includes repeated values, outliers, missing value processing, normalization, and the like, and the present application is not limited to this specific one.

Wherein, the repeated value processing means deleting a plurality of pieces of data repeatedly appearing at the same time in the data; outlier processing refers to deleting obvious outliers in the data; the missing value processing refers to the completion of missing moment data in the historical data; normalization refers to normalizing training data using a log1p algorithm, and normalization refers to using a minimum maximum scale.

The normalized and preprocessed training data is input into a preset LSTM model, so that the training precision of the model can be further improved, and the service life of the target network element can be predicted better.

For example, as shown in fig. 2, a schematic diagram of a training process of a network element lifetime prediction model according to the present application is shown.

In one implementation, determining a loss value of a preset long-term and short-term memory network model by comparing a life to be verified with a labeling result of training data includes:

And calculating an average square error between the service life to be verified and the labeling result of the training data, and taking the average square error as a loss value of a preset long-and-short-term memory network model.

That is, in the present application, an MSE (mean-square error) loss function may be selected as the loss function during training of the preset LSTM model. Further, an Adam optimization algorithm may be selected to converge on the preset LSTM model, which is not limited in the present application.

In step S13, after the current performance data and the current alarm number of the target network element are obtained, the current performance data and the current alarm number are input into the network element lifetime prediction model, and the remaining available time of the target network element from the current time is predicted, so as to obtain the predicted lifetime of the target network element device.

In the step, in the actual production process, after the current performance data and the current alarm number of the target network element are obtained, the current performance data and the current alarm number can be input into a network element life prediction model, the residual available time of the target network element from the current time is predicted to obtain the predicted life of the target network element equipment, and then the predicted life is substituted into the current month depreciation cost mathematical model by combining with expert service experience to calculate the current month depreciation cost.

In one implementation, inputting current performance data and a current alarm number into a network element lifetime prediction model, predicting a remaining available time length of a target network element from a current time to obtain a predicted lifetime of target network element equipment, including:

Inputting the current performance data and the current alarm number into a network element life prediction model, and splicing the current performance data and the current alarm number by an input layer to obtain input layer data; modeling the time dependency relationship in the input layer data by the long-term and short-term memory network layer structure to obtain characteristic data; and the full connection layer converts the feature space dimension of the feature data and carries out fitting processing to obtain the predicted service life of the target network element equipment.

That is, the preset LSTM model includes a multi-layer structure including an input layer, a long-short-term memory network layer structure, a full-connection layer, and the like.

The LSTM with the long-term memory network layer structure can effectively solve the long-sequence problem by introducing the concepts of memory cells, input gates, output gates and forgetting gates. The memory cell is responsible for storing important information, the input gate decides not to write current input information into the memory cell, the forget gate decides not to forget information in the memory cell, and the output gate decides not to take the information of the memory cell as current output. The control of these gates can effectively capture the important long-term dependencies in the sequence and can solve the gradient problem.

Therefore, the LSTM is more suitable for processing and predicting important events with longer intervals in a time sequence, can learn data characteristics of network element work more fully, can obtain a finer modeling effect than manual modeling, and is beneficial to improving accuracy of network element service life prediction results.

In one implementation, the full connection layer performs conversion and overfitting processing on feature space dimensions of feature data to obtain a predicted lifetime of the target network element device, including:

The full connection layer converts and overfits the feature space dimension of the feature data to obtain the predicted alarm number of the target network element equipment from the current time; comparing the predicted alarm number with an alarm number threshold corresponding to the target network element, and determining an unavailable time point of the target network element equipment; and calculating the difference value between the unavailable time point and the current time to serve as the predicted service life of the target network element equipment.

That is, after obtaining the predicted alert number of the predicted future target network element, the alert number threshold corresponding to the target network element counted by the expert may be further compared with the predicted alert number of the target network element, so as to determine the unavailable time point of the target network element in the future, and further, the time difference from the unavailable time point to the current time point is taken as the predicted lifetime of the target network element.

As can be seen from the description in the foregoing steps, the threshold value of the alarm number corresponding to the target network element is an average value of the alarm numbers of the network element devices of the same type as the target network element when the network element devices are unavailable, or may be an empirical value which is agreed in advance, which is not performed by the present application.

For example, as shown in fig. 3, a schematic diagram of steps of an implementation of the present application is shown.

The model structure of the network element life prediction model is shown in fig. 4, the input current performance data and the current alarm number feature are spliced through Concat layers (input layers), the time dependency relationship in the data is modeled through a multi-layer LSTM structure, then the feature space dimension is converted through a Dense layer (full connection layer), and further, the overfitting is prevented and the convergence speed is increased through the Batch Norm and Dropout structure, so that the predicted alarm number of the predicted target network element device from the current time is obtained.

Furthermore, by comparing the predicted alarm number with the alarm number threshold corresponding to the target network element, the unavailable time point of the target network element equipment can be determined, and then the difference between the unavailable time point and the current time is calculated and used as the predicted service life of the target network element equipment.

In step S14, the current depreciation cost of the target network element is calculated according to the predicted lifetime.

In this step, a mathematical model of the current depreciation cost is built, and the predicted lifetime of the target network element is introduced into the calculation of the depreciation cost, so that the actual situation of the service can be more attached to the calculation of the depreciation cost according to the time proportion, and the service lifetime of the target network element can be more accurately predicted and the depreciation cost of each month can be reduced compared with the conservative fixed service life preset in the existing depreciation cost.

Therefore, the device for precisely calculating the resource depreciation cost of the 5G base station provided by the invention predicts the service life of each network element device more precisely, realizes a network element service life prediction model driven by data, has high reliability and is more close to actual production data, greatly reduces modeling difficulty of developers, can quickly iterate the model according to actual production feedback, and improves the prediction accuracy of the model.

In one implementation, calculating a current depreciation cost of a target network element based on a predicted lifetime includes:

Acquiring a fixed asset original value of a target network element; and taking the ratio of the original value of the fixed asset to the predicted life as the current depreciation cost of the target network element.

That is, in the current implementation, the current depreciation mathematical model involves several concepts: the estimated monthly depreciation cost, the fixed asset original value, the network element estimated life (years), the current month net value. The fixed asset original value represents the purchase original price of the fixed asset; the predicted service life (year) of the network element is the service life of the network element predicted by the network element service life prediction model; the estimated monthly depreciation cost is the current depreciation cost.

The fixed asset original value and the network element predicted life (year) are input into the depreciation cost mathematical model, and the depreciation rate in the current month is output from the depreciation cost mathematical model.

Then, the estimated monthly depreciation cost=fixed asset raw value/(12 x network element estimated lifetime).

In one implementation, a ratio of a fixed asset primary value to a predicted lifetime is used as a current depreciation cost of a target network element, comprising:

Taking the ratio of the original value of the fixed asset to the predicted life as the candidate current depreciation cost of the target network element; acquiring historical depreciation cost of a target network element; and calculating a difference value between the original value of the fixed asset and the historical depreciation cost, and taking the minimum value between the difference value and the candidate current depreciation cost as the current depreciation cost of the target network element.

That is, the current net value of the target network element, i.e., the difference between the fixed asset original value and the historical depreciation cost, may be calculated prior to the final determination of the current depreciation cost.

I.e. when net value = fixed asset original value- Σmonth depreciation.

Then, when the net value < the estimated monthly depreciation cost, then the monthly depreciation cost = the net value;

the current net value > the predicted month depreciation cost, then the current month depreciation cost = predicted month depreciation cost.

Therefore, the actual service life of the equipment can be accurately predicted, the depreciation cost of the equipment per month is reduced, and the operation cost is further reduced. And the health condition of the network element can be evaluated more accurately, the basis for equipment replacement is provided for daily operation and maintenance of the base station, the manual inspection cost is reduced, and considerable economic benefits are realized.

The method and the system can be suitable for evaluating the benefit value of the 5G base station, guide operators to optimize targeted construction sites and site selection, greatly improve the accuracy of site benefit value evaluation, provide data support for the operators to construct the 5G sites, reduce the construction investment of low-benefit sites and improve the site construction investment return rate.

From the above, it can be seen that, according to the technical scheme provided by the embodiment of the present application, through the network element lifetime prediction model obtained by training, the lifetime of the target network element is predicted based on the current alarm number of the current performance data set, and further, based on the remaining lifetime of the target network element, the corresponding depreciation cost is calculated, and compared with the method of directly calculating the depreciation cost according to the use time in the related art, the remaining service lifetime of the target network element can be predicted more accurately, the obtained current depreciation cost can also more conform to the actual use situation of the target network element, so that the accuracy of cost estimation of the target network element is higher, the depreciation cost of each month is further reduced, and the operation cost is reduced.

According to the network element equipment depreciation prediction method provided by the embodiment of the application, the execution main body can be the network element equipment depreciation prediction device. In the embodiment of the application, a method for executing terminal access by using a network element equipment depreciation prediction device is taken as an example, and the device for the network element equipment depreciation prediction method provided by the embodiment of the application is described.

Fig. 5 is a block diagram of a network element device depreciation prediction apparatus according to an exemplary embodiment, including:

an obtaining module 201, configured to obtain historical performance data of a target network element and a corresponding historical alarm number;

The training module 202 is configured to train a preset long-short-period memory network model based on the historical performance data and the historical alarm number, and iteratively adjust model parameters of the preset long-short-period memory network model to obtain a network element lifetime prediction model;

A prediction module 203, configured to obtain current performance data and a current alarm number of the target network element, and then input the current performance data and the current alarm number into the network element lifetime prediction model, and predict a remaining available time span of the target network element from a current time to obtain a predicted lifetime of the target network element device;

a calculating module 204, configured to calculate a current depreciation cost of the target network element according to the predicted lifetime.

Optionally, the training module 202 is configured to:

Acquiring type information of the target network element, and determining an alarm number threshold corresponding to the target network element based on the type information;

Marking the historical alarm number and the corresponding historical performance data as negative samples when the historical alarm number is larger than or equal to the alarm number threshold value, and marking the historical alarm number and the corresponding historical performance data as positive samples when the historical alarm number is smaller than the alarm number threshold value, so as to obtain training data;

and training a preset long-short-period memory network model based on the training data, and iteratively adjusting model parameters of the preset long-short-period memory network model to obtain a network element life prediction model.

Optionally, the training module 202 is configured to:

inputting the training data into a preset long-term and short-term memory network model, and predicting the remaining available time length of the target network element from the corresponding time of the training data to obtain the service life of the target network element to be verified;

Comparing the service life to be verified with the labeling result of the training data, and determining the loss value of the preset long-period and short-period memory network model;

And under the condition that the loss value is larger than a preset loss threshold value, iteratively adjusting the model parameters of the preset long-short-period memory network model until the loss value is smaller than or equal to the preset loss threshold value, wherein the adjusted preset long-short-period memory network model is used as a network element life prediction model.

Optionally, the training module 203 is configured to:

And carrying out normalization processing and normalization processing on the training data, inputting the processed training data into a preset long-term and short-term memory network model, and predicting the residual available time length of the target network element from the corresponding time of the training data to obtain the service life to be verified of the target network element.

Optionally, the training module 203 is configured to:

And calculating an average square error between the service life to be verified and the labeling result of the training data, and taking the average square error as a loss value of the preset long-short-term memory network model.

Optionally, the prediction module 203 is configured to:

Inputting the current performance data and the current alarm number into the network element life prediction model, and splicing the current performance data and the current alarm number by an input layer to obtain input layer data;

Modeling the time dependency relationship in the input layer data by using a long-term and short-term memory network layer structure to obtain characteristic data;

and the full connection layer converts and overfits the feature space dimension of the feature data to obtain the predicted service life of the target network element equipment.

Optionally, the prediction module 203 is configured to:

the full connection layer converts and overfits the feature space dimension of the feature data to obtain the predicted alarm number of the target network element equipment from the current time;

comparing the predicted alarm number with an alarm number threshold corresponding to the target network element, and determining an unavailable time point of the target network element equipment;

And calculating the difference value between the unavailable time point and the current time to serve as the predicted service life of the target network element equipment.

Optionally, the computing module 204 is configured to:

Acquiring a fixed asset original value of the target network element;

and taking the ratio of the original value of the fixed asset to the predicted life as the current depreciation cost of the target network element.

Optionally, the computing module 204 is configured to:

Taking the ratio of the fixed asset original value to the predicted life as the candidate current depreciation cost of the target network element;

Acquiring the historical depreciation cost of the target network element;

And calculating a difference value between the original value of the fixed asset and the historical depreciation cost, and taking the minimum value of the difference value and the candidate current depreciation cost as the current depreciation cost of the target network element.

From the above, it can be seen that, according to the technical scheme provided by the embodiment of the present application, through the network element lifetime prediction model obtained by training, the lifetime of the target network element is predicted based on the current alarm number of the current performance data set, and further, based on the remaining lifetime of the target network element, the corresponding depreciation cost is calculated, and compared with the method of directly calculating the depreciation cost according to the use time in the related art, the remaining service lifetime of the target network element can be predicted more accurately, the obtained current depreciation cost can also more conform to the actual use situation of the target network element, so that the accuracy of cost estimation of the target network element is higher, the depreciation cost of each month is further reduced, and the operation cost is reduced.

According to the network element equipment depreciation prediction method provided by the embodiment of the application, the execution main body can be a terminal access terminal. In the embodiment of the application, a method for a terminal access terminal to execute terminal access is taken as an example, and a device of the network element equipment depreciation prediction method provided by the embodiment of the application is described.

The network element equipment depreciation prediction device in the embodiment of the application can be electronic equipment, and can also be a component in the electronic equipment, such as an integrated circuit or a chip. The electronic device may be a terminal, or may be other devices than a terminal. The electronic device may be a Mobile phone, a tablet computer, a notebook computer, a palm computer, a vehicle-mounted electronic device, a Mobile internet appliance (Mobile INTERNET DEVICE, MID), an augmented reality (augmented reality, AR)/Virtual Reality (VR) device, a robot, a wearable device, an ultra-Mobile personal computer (UMPC), a netbook or a Personal Digital Assistant (PDA), etc., and may also be a server, a network attached storage (Network Attached Storage, NAS), a personal computer (personal computer, PC), a Television (TV), a teller machine, a self-service machine, etc., which are not particularly limited in the embodiments of the present application.

The network element equipment depreciation prediction device provided by the embodiment of the present application can implement each process implemented by the embodiments of the methods of fig. 1 to fig. 4, and in order to avoid repetition, a description thereof is omitted.

Optionally, as shown in fig. 6, the embodiment of the present application further provides an electronic device 500, including a processor 501 and a memory 502, where the memory 502 stores a program or an instruction that can be executed on the processor 501, and the program or the instruction implements each step of the above embodiment of the method for predicting depreciation of a network element device when executed by the processor 501, and can achieve the same technical effect, so that repetition is avoided, and no further description is given here.

The electronic device in the embodiment of the application includes the mobile electronic device and the non-mobile electronic device.

Fig. 7 is a schematic diagram of a hardware structure of an electronic device implementing an embodiment of the present application.

The electronic device 1000 includes, but is not limited to: radio frequency unit 1001, network module 1002, audio output unit 1003, input unit 1004, sensor 1005, display unit 1006, user input unit 1007, interface unit 1008, memory 1009, and processor 1010.

Those skilled in the art will appreciate that the electronic device 1000 may also include a power source (e.g., a battery) for powering the various components, which may be logically connected to the processor 1010 by a power management system to perform functions such as managing charge, discharge, and power consumption by the power management system. The electronic device structure shown in fig. 7 does not constitute a limitation of the electronic device, and the electronic device may include more or less components than shown, or may combine certain components, or may be arranged in different components, which are not described in detail herein.

From the above, it can be seen that, according to the technical scheme provided by the embodiment of the present application, through the network element lifetime prediction model obtained by training, the lifetime of the target network element is predicted based on the current alarm number of the current performance data set, and further, based on the remaining lifetime of the target network element, the corresponding depreciation cost is calculated, and compared with the method of directly calculating the depreciation cost according to the use time in the related art, the remaining service lifetime of the target network element can be predicted more accurately, the obtained current depreciation cost can also more conform to the actual use situation of the target network element, so that the accuracy of cost estimation of the target network element is higher, the depreciation cost of each month is further reduced, and the operation cost is reduced.

It should be appreciated that in embodiments of the present application, the input unit 1004 may include a graphics processor (Graphics Processing Unit, GPU) 10041 and a microphone 10042, where the graphics processor 10041 processes image data of still pictures or video obtained by an image capturing device (e.g., a camera) in a video capturing mode or an image capturing mode. The display unit 1006 may include a display panel 10061, and the display panel 10061 may be configured in the form of a liquid crystal display, an organic light emitting diode, or the like. The user input unit 1007 includes at least one of a touch panel 10071 and other input devices 10072. The touch panel 10071 is also referred to as a touch screen. The touch panel 10071 can include two portions, a touch detection device and a touch controller. Other input devices 10072 may include, but are not limited to, a physical keyboard, function keys (e.g., volume control keys, switch keys, etc.), a trackball, a mouse, a joystick, and so forth, which are not described in detail herein.

The memory 1009 may be used to store software programs as well as various data. The memory 1009 may mainly include a first memory area storing programs or instructions and a second memory area storing data, wherein the first memory area may store an operating system, application programs or instructions (such as a sound playing function, an image playing function, etc.) required for at least one function, and the like. Further, the memory 1009 may include volatile memory or nonvolatile memory, or the memory 1009 may include both volatile and nonvolatile memory. The nonvolatile Memory may be a Read-Only Memory (ROM), a Programmable ROM (PROM), an Erasable PROM (EPROM), an Electrically Erasable EPROM (EEPROM), or a flash Memory. The volatile memory may be random access memory (Random Access Memory, RAM), static random access memory (STATIC RAM, SRAM), dynamic random access memory (DYNAMIC RAM, DRAM), synchronous Dynamic Random Access Memory (SDRAM), double data rate Synchronous dynamic random access memory (Double DATA RATE SDRAM, DDRSDRAM), enhanced Synchronous dynamic random access memory (ENHANCED SDRAM, ESDRAM), synchronous link dynamic random access memory (SYNCH LINK DRAM, SLDRAM), and Direct random access memory (DRRAM). Memory 109 in embodiments of the present application includes, but is not limited to, these and any other suitable types of memory.

The processor 1010 may include one or more processing units; optionally, the processor 1010 integrates an application processor that primarily processes operations involving an operating system, user interface, application programs, and the like, and a modem processor that primarily processes wireless communication signals, such as a baseband processor. It will be appreciated that the modem processor described above may not be integrated into the processor 1010.

The embodiment of the application also provides a readable storage medium, and the readable storage medium stores a program or an instruction, which when executed by a processor, implements each process of the network element equipment depreciation prediction method embodiment, and can achieve the same technical effect, so that repetition is avoided, and no further description is provided herein.

Wherein the processor is a processor in the electronic device described in the above embodiment. The readable storage medium includes computer readable storage medium such as computer readable memory ROM, random access memory RAM, magnetic or optical disk, etc.

The embodiment of the application further provides a chip, the chip comprises a processor and a communication interface, the communication interface is coupled with the processor, the processor is used for running a program or instructions, the processes of the network element equipment depreciation prediction method embodiment can be realized, the same technical effects can be achieved, and the repetition is avoided, and the description is omitted here.

It should be understood that the chips referred to in the embodiments of the present application may also be referred to as system-on-chip chips, chip systems, or system-on-chip chips, etc.

An embodiment of the present application provides a computer program product, which is stored in a storage medium, and the program product is executed by at least one processor to implement the respective processes of the embodiment of the network element device depreciation prediction method, and achieve the same technical effects, so that repetition is avoided, and a detailed description is omitted herein.

It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element. Furthermore, it should be noted that the scope of the methods and apparatus in the embodiments of the present application is not limited to performing the functions in the order shown or discussed, but may also include performing the functions in a substantially simultaneous manner or in an opposite order depending on the functions involved, e.g., the described methods may be performed in an order different from that described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

From the above description of the embodiments, it will be clear to those skilled in the art that the above-described embodiment method may be implemented by means of software plus a necessary general hardware platform, but of course may also be implemented by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art in the form of a computer software product stored in a storage medium (e.g. ROM/RAM, magnetic disk, optical disk) comprising instructions for causing a terminal (which may be a mobile phone, a computer, a server, or a network device, etc.) to perform the method according to the embodiments of the present application.

The embodiments of the present application have been described above with reference to the accompanying drawings, but the present application is not limited to the above-described embodiments, which are merely illustrative and not restrictive, and many forms may be made by those having ordinary skill in the art without departing from the spirit of the present application and the scope of the claims, which are to be protected by the present application.

Claims (12)

1. A method for predicting depreciation of network element equipment, comprising:

acquiring historical performance data and corresponding historical alarm numbers of a target network element;

Training a preset long-short-period memory network model based on the historical performance data and the historical alarm number, and iteratively adjusting model parameters of the preset long-short-period memory network model to obtain a network element life prediction model;

After the current performance data and the current alarm number of the target network element are obtained, the current performance data and the current alarm number are input into the network element life prediction model, and the residual available time length of the target network element from the current time is predicted to obtain the predicted life of the target network element equipment;

and calculating the current depreciation cost of the target network element according to the predicted service life.

2. The network element equipment depreciation prediction method according to claim 1, wherein training the preset long-short-period memory network model based on the historical performance data and the historical alarm number, iteratively adjusting model parameters of the preset long-short-period memory network model, and obtaining a network element lifetime prediction model comprises:

Acquiring type information of the target network element, and determining an alarm number threshold corresponding to the target network element based on the type information;

Marking the historical alarm number and the corresponding historical performance data as negative samples when the historical alarm number is larger than or equal to the alarm number threshold value, and marking the historical alarm number and the corresponding historical performance data as positive samples when the historical alarm number is smaller than the alarm number threshold value, so as to obtain training data;

and training a preset long-short-period memory network model based on the training data, and iteratively adjusting model parameters of the preset long-short-period memory network model to obtain a network element life prediction model.

3. The network element equipment depreciation prediction method according to claim 2, wherein the training the preset long-short-term memory network model based on the training data, iteratively adjusting model parameters of the preset long-short-term memory network model, and obtaining a network element lifetime prediction model, includes:

inputting the training data into a preset long-term and short-term memory network model, and predicting the remaining available time length of the target network element from the corresponding time of the training data to obtain the service life of the target network element to be verified;

Comparing the service life to be verified with the labeling result of the training data, and determining the loss value of the preset long-period and short-period memory network model;

And under the condition that the loss value is larger than a preset loss threshold value, iteratively adjusting the model parameters of the preset long-short-period memory network model until the loss value is smaller than or equal to the preset loss threshold value, wherein the adjusted preset long-short-period memory network model is used as a network element life prediction model.

4. The network element equipment depreciation prediction method according to claim 3, wherein the inputting the training data into a preset long-term memory network model predicts the remaining available time length of the target network element from the training data corresponding time to obtain the service life to be verified of the target network element, and comprises:

And carrying out normalization processing and normalization processing on the training data, inputting the processed training data into a preset long-term and short-term memory network model, and predicting the residual available time length of the target network element from the corresponding time of the training data to obtain the service life to be verified of the target network element.

5. The network element equipment depreciation prediction method according to claim 3, wherein the comparing the service life to be verified with the labeling result of the training data to determine the loss value of the preset long-term and short-term memory network model comprises:

And calculating an average square error between the service life to be verified and the labeling result of the training data, and taking the average square error as a loss value of the preset long-short-term memory network model.

6. The network element equipment depreciation prediction method according to claim 1, wherein the inputting the current performance data and the current alarm number into the network element lifetime prediction model predicts the remaining available time length of the target network element from the current time to obtain the predicted lifetime of the target network element equipment includes:

Inputting the current performance data and the current alarm number into the network element life prediction model, and splicing the current performance data and the current alarm number by an input layer to obtain input layer data;

Modeling the time dependency relationship in the input layer data by using a long-term and short-term memory network layer structure to obtain characteristic data;

and the full connection layer converts and overfits the feature space dimension of the feature data to obtain the predicted service life of the target network element equipment.

7. The network element equipment depreciation prediction method according to claim 6, wherein the converting and fitting the feature space dimension of the feature data by the full connection layer to obtain the predicted lifetime of the target network element equipment includes:

the full connection layer converts and overfits the feature space dimension of the feature data to obtain the predicted alarm number of the target network element equipment from the current time;

comparing the predicted alarm number with an alarm number threshold corresponding to the target network element, and determining an unavailable time point of the target network element equipment;

And calculating the difference value between the unavailable time point and the current time to serve as the predicted service life of the target network element equipment.

8. The network element equipment depreciation prediction method according to claim 1, wherein calculating the current depreciation cost of the target network element according to the predicted lifetime comprises:

Acquiring a fixed asset original value of the target network element;

and taking the ratio of the original value of the fixed asset to the predicted life as the current depreciation cost of the target network element.

9. The network element equipment depreciation prediction method according to claim 8, wherein the step of using the ratio of the fixed asset primary value and the predicted lifetime as the current depreciation cost of the target network element comprises:

Taking the ratio of the fixed asset original value to the predicted life as the candidate current depreciation cost of the target network element;

Acquiring the historical depreciation cost of the target network element;

And calculating a difference value between the original value of the fixed asset and the historical depreciation cost, and taking the minimum value of the difference value and the candidate current depreciation cost as the current depreciation cost of the target network element.

10. A network element device depreciation prediction apparatus, comprising:

the acquisition module is used for acquiring the historical performance data of the target network element and the corresponding historical alarm number;

The training module is used for training a preset long-short-period memory network model based on the historical performance data and the historical alarm number, and iteratively adjusting model parameters of the preset long-short-period memory network model to obtain a network element service life prediction model;

The prediction module is used for inputting the current performance data and the current alarm number into the network element life prediction model after acquiring the current performance data and the current alarm number of the target network element, and predicting the residual available time length of the target network element from the current time to obtain the predicted life of the target network element equipment;

And the calculating module is used for calculating the current depreciation cost of the target network element according to the predicted service life.

11. An electronic device, comprising:

a processor;

a memory for storing the processor-executable instructions;

wherein the processor is configured to execute the instructions to implement the network element device depreciation prediction method according to any of claims 1 to 9.

12. A readable storage medium, wherein a program or instructions is stored on the readable storage medium, which when executed by a processor, implements the steps of the network element device depreciation prediction method according to any of claims 1-9.