Regional project cost analysis and prediction method and system based on neural network

Abstract

The invention provides a regional cost analysis and prediction method and system based on a neural network. The method comprises the following steps: constructing a neural network model; extracting data from a historical cost data information base as a model input layer as an independent variable to construct a neural network model, and setting an output layer dependent variable at the same time; the historical cost data information base comprises a regional economic condition part, a consumption level part, a market demand change part and a competitor dynamic variable part; training the constructed neural network model, dividing the data into a training set and a test set, and performing model training through the training set to obtain the required neural network model; and carrying out real-time cost analysis and prediction based on the constructed neural network model. According to the method, the regional project cost can be quickly analyzed and predicted, and managers can be better assisted in effective budget planning and control.

Description

Technical Field

[0001] The present invention belongs to the technical field of data analysis, and in particular relates to a method and system for analyzing and predicting regional project costs based on a neural network. Background Art

[0002] In the current field of photovoltaic engineering management and construction, project cost management is a key link. Traditional cost calculation and data management methods often rely on manual calculation operations, which can easily lead to inaccurate data statistics, low calculation efficiency, and lack of real-time update capabilities in budget control. In addition, existing data management systems can usually only integrate and analyze a single area independently, and cannot support comprehensive analysis of multiple regions and multiple projects. This will greatly limit the support for enterprise managers to obtain effective decision-making, making the management of regional project cost data inefficient. With the increasing requirements for cost control and efficiency improvement in the industry, how to effectively integrate the data of project cost management in each region to better analyze and visualize cost data and provide better decision-making basis for enterprises is a technical problem that needs to be solved at present. Summary of the invention

[0003] The invention provides a method and system for analyzing and predicting regional project costs based on a neural network.

[0004] The purpose of the present invention is achieved through the following technical solutions: A method for regional cost analysis and prediction based on a neural network, the method comprising the following steps: S1. Constructing a neural network model; extracting required data from a historical cost data information base as the independent variables required by the model input layer to construct a neural network model, and setting the dependent variables of the output layer; the historical cost data information base includes a regional economic situation part, a consumption level part, a market demand change part, and a competitor dynamic variable part; S2, training the neural network model constructed in S1, dividing the data in S1 into a training set and a test set, and performing model training on the training set to obtain the required neural network model; and during the construction process, a multi-layer perceptron is used to perform activation functions to introduce nonlinear factors to increase the nonlinear fitting ability of the neural network; S3. Analyze and predict real-time costs based on the constructed neural network model.

[0005] Preferably, the data of the regional economic situation in S1 includes regional GDP data and unemployment rate data; The data of the consumption level section include per capita disposable income and per capita consumption index; The data on market demand changes include demand fluctuation rates over the years; The data of competitor dynamic variables include market share change rate and frequency of new product launches.

[0006] Preferably, the dependent variables in S1 include development service fee data, material cost data and warranty deposit data.

[0007] Preferably, in the process of training the constructed neural network model in S2, in order to minimize the loss function, the model parameters are gradually adjusted through an optimization algorithm, and the optimization algorithm includes an Adam optimization algorithm or an SGD optimization algorithm.

[0008] Preferably, the real-time costs in S3 include development service fees, material fees and warranty fees. Preferably, a system for regional cost analysis and prediction method based on neural network includes: A historical cost data acquisition module is used to acquire historical cost data; A data storage module, electrically connected to the historical expense data module, for storing the acquired historical expense data; A multi-dimensional data prediction module is electrically connected to the data storage module and is used to predict the data in the storage module. Dynamic data analysis module, which analyzes the data in the data storage module; The visual display module is electrically connected to the data storage module and displays the data in the data storage module and the data analyzed by the dynamic data analysis module.

[0009] Preferably, a computer-readable storage medium stores a computer program, which is executed by a processor to implement the steps of any of the above methods.

[0010] The beneficial effects of the present invention are reflected in that the present invention can quickly analyze and predict regional project costs, and better assist managers in effective budget planning and control. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying any creative work.

[0012] Figure 1 : A schematic diagram of the relationship between the various layers in the model constructed based on the neural network of the present invention. DETAILED DESCRIPTION

[0013] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the following Figure 1It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0014] The present invention proposes a regional cost analysis and prediction system based on a neural network and a method thereof, wherein the regional cost analysis and prediction method system based on a neural network comprises: The historical cost data acquisition module is used to acquire historical cost data; the historical cost data includes regional economic situation data, consumption level data, market demand change data, and competitor dynamic variable data. Furthermore, the regional economic situation data includes regional GDP data and unemployment rate data; the consumption level data includes per capita disposable income and per capita consumption index; the market demand change data includes demand volatility over the years, and the competitor dynamic variable data includes market share change rate and new product launch frequency. Historical data can usually be acquired through targeted market surveys, public industry or financial reports, and social media and news monitoring.

[0015] A data storage module, electrically connected to the historical expense data module, for storing the acquired historical expense data; A multi-dimensional data prediction module, which is based on a neural network and is electrically connected to the data storage module, and is used to predict the data in the storage module and to periodically perform breakpoint continuous training on the constructed model; Dynamic data analysis module, which analyzes the data in the data storage module; The visualization display module is electrically connected to the data storage module to display the data in the data storage module, wherein the visualization display module can further display the data analyzed by the dynamic data analysis module.

[0016] The present invention also discloses a method for analyzing and predicting a regional cost analysis and prediction system based on a neural network, comprising the following steps: S1. Construct a neural network model; use the historical cost data acquisition module to extract the required data from the historical cost data information library as the independent variables required by the model input layer to construct the neural network model, and set the dependent variables of the output layer; the historical cost data information library includes the regional economic situation part, the consumption level part, the market demand change part, and the competitor dynamic variable part; combine Figure 1 As shown, the neural network model includes an input layer, a hidden layer and an output layer. During the construction process, a multi-layer perceptron is used to introduce nonlinear factors into the activation function to increase the nonlinear fitting ability of the neural network; in this embodiment, the dependent variables include development service fee data, material fee data and warranty deposit data.

[0017] S2. Train the neural network model constructed in S1, divide the data in S1 into a training set and a test set, and perform model training through the training set to obtain the required neural network model; in the process of training the constructed neural network model, in order to minimize the loss function, gradually adjust the model parameters through an optimization algorithm, and the optimization algorithm includes an Adam optimization algorithm or an SGD optimization algorithm.

[0018] The model training method described in the present invention comprises the following steps: S21, forward propagation: pass the input data through the neural network, obtain the output value, and compare it with the actual label to get the error; S22, back propagation: According to the error in S21, the influence of each parameter on the error is obtained layer by layer according to the chain rule, and the parameters are updated; The update parameters are updated based on the gradient information obtained by back propagation, using an optimization algorithm (such as gradient descent) to update the weights and bias parameters in the network; S23, repeated training: Repeat the above steps, i.e. forward propagation, back propagation and parameter update, until the specified stopping condition is reached (such as reaching the maximum number of training times or the error is less than the set threshold). The model obtained at this time is the optimized model.

[0019] S3: Analyze and predict the real-time costs based on the constructed neural network model. The real-time costs in S3 include development service fees, material fees and warranty fees.

[0020] When in use, the system collects one or more data such as GDP, unemployment rate, per capita income, consumption index, demand volatility, market change rate, new product frequency, etc. in the input layer of the system, and obtains the predicted values ​​of development service fee, material fee and warranty after analysis and prediction. The development service fee, material fee and warranty are the output y of the model, and the true value is , output ; The loss function is ;in, is the weight parameter matrix of the first layer, is the bias, relu is the activation function; is the weight parameter matrix of the second layer, For bias.

[0021] Since changes in different data in the input layer will affect changes in different dependent variables in the final output layer, it can better assist enterprise managers in judging market changes through the relationship between independent variables and dependent variables, thereby improving decision-making efficiency and operational results.

[0022] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A regional cost analysis and prediction method based on a neural network, characterized in that: The method comprises the following steps: S1. Constructing a neural network model; extracting required data from a historical cost data information base as the independent variables required by the model input layer to construct a neural network model, and setting the dependent variables of the output layer; the historical cost data information base includes a regional economic situation part, a consumption level part, a market demand change part, and a competitor dynamic variable part; S2, training the neural network model constructed in S1, dividing the data in S1 into a training set and a test set, and performing model training on the training set to obtain the required neural network model; and during the construction process, a multi-layer perceptron is used to perform activation functions to introduce nonlinear factors to increase the nonlinear fitting ability of the neural network; S3. Analyze and predict real-time costs based on the constructed neural network model.

2. A method for regional cost analysis and prediction based on a neural network as claimed in claim 1, characterized in that: The data of the regional economic situation in S1 include regional GDP data and unemployment rate data; The data of the consumption level section include per capita disposable income and per capita consumption index; The data on market demand changes include demand fluctuation rates over the years; The data of competitor dynamic variables include market share change rate and frequency of new product launches.

3. A method for regional cost analysis and prediction based on a neural network as claimed in claim 1, characterized in that: The dependent variables in S1 include development service fee data, material cost data and warranty deposit data.

4. A method for regional cost analysis and prediction based on a neural network as claimed in claim 1, characterized in that: In the process of training the constructed neural network model in S2, in order to minimize the loss function, the model parameters are gradually adjusted through an optimization algorithm, and the optimization algorithm includes an Adam optimization algorithm or an SGD optimization algorithm.

5. A method for regional cost analysis and prediction based on neural network as claimed in claim 1, characterized in that: The real-time fees in S3 include development service fees, material fees and warranty fees.

6. A system for regional cost analysis and prediction method based on neural network, characterized in that: include, A historical cost data acquisition module is used to acquire historical cost data; A data storage module, electrically connected to the historical expense data module, for storing the acquired historical expense data; A multi-dimensional data prediction module, electrically connected to the data storage module, for predicting the data in the storage module; Dynamic data analysis module, which analyzes the data in the data storage module; The visual display module is electrically connected to the data storage module and displays the data in the data storage module and the data analyzed by the dynamic data analysis module.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the steps of the method according to any one of claims 1 to 5.

Images