CN110807490A - An intelligent prediction method of transmission line engineering cost based on single base tower - Google Patents

Abstract

The method for intelligently predicting the cost of a transmission line project based on a single base tower provided by the present application includes the following steps: combining the empirical parameters of the least squares support vector machine model to construct an empirical parameter prediction model; Optimize the parameters of the particle swarm optimization least squares support vector machine to obtain the particle swarm optimization least squares support vector machine prediction model; use the original data as input variables, input them into the optimized vector machine prediction model, and train the vector machine prediction model to obtain Predicted cost of a single base tower. The method provided in this application uses the principal component analysis method to reduce the dimensionality of the indicators, innovatively introduces the particle swarm algorithm to optimize the parameters of the least squares support vector machine model to obtain the optimal parameters, and import the obtained principal component data into the Training and prediction in the empirical parameter prediction model and the optimized parameter prediction model can improve the accuracy of cost prediction and improve the level of refinement of cost management.

Description

An intelligent prediction method of transmission line engineering cost based on single base tower

technical field

The present application relates to the technical field of electric power engineering construction, and in particular to an intelligent prediction method of transmission line engineering cost based on a single base tower.

Background technique

In recent years, with the continuous expansion of the investment scale and construction scale of my country's power grid construction projects, strengthening the cost management of power grid projects and realizing the lean management of power grid investment have become an important research topic faced by power grid enterprises in the development process. By effectively calculating the actual cost of each base tower, the construction of a single base tower model pricing framework system and pricing method can improve the cost control level of power grid construction projects such as feasibility study estimates, preliminary design estimates, construction drawing budgets, and completion settlement. It is of great significance to manage the cost of the whole process of construction and improve the level of lean management of power grid construction investment.

The new round of power system reform centered on the reform of transmission and distribution prices has completely changed the profit model of State Grid Corporation. Therefore, as the core work of controlling the company's construction cost, power grid cost management must follow the development principles of high quality and high efficiency. In the future, we must focus on the control of power grid investment cost and strengthen the precise investment and control of infrastructure. In order to meet the strategic development needs of the State Grid Corporation of China, it is necessary to innovate the project construction and management model, and to explore new ideas for power grid project cost management in the new era. Taking serving the power grid construction as the starting point, the ability and level of project cost management and control must be comprehensively improved.

As an important production, operation and maintenance carrier of the company, overhead line project has many factors affecting the cost. Differences in topography, geology, meteorology, transportation conditions and other factors will cause large cost differences. For a long time, the project budget, completion settlement and financial final accounts of overhead transmission lines have been priced according to the entire line, and the project construction management and asset collection have not yet been refined. The actual cost level and cost accuracy of the base tower need to be further improved. Therefore, it is necessary to innovate the pricing model, strengthen the cost management of power grid engineering construction, reasonably control the project investment, and improve the efficiency of power grid construction.

SUMMARY OF THE INVENTION

The present application provides an intelligent forecasting method for transmission line engineering cost based on a single base tower, which is used to solve the problems of unscientific forecasting methods and low forecasting accuracy in existing forecasting methods.

The technical solutions adopted by the present application to solve the above-mentioned technical problems are as follows:

An intelligent prediction method for the construction cost of a transmission line based on a single base tower, the method comprises the following steps:

Combine the empirical parameters of the least squares support vector machine model to build the empirical parameter prediction model;

Using the particle swarm optimization algorithm to optimize the parameters in the prediction model to obtain the particle swarm optimization least squares support vector machine prediction model;

The original data is used as an input variable to be input into the optimized vector machine prediction model, and the vector machine prediction model is trained to obtain the cost prediction value of a single base tower.

Optionally, before the original data is input to the optimized vector machine prediction model as an input variable, the method further includes:

Collect historical data of the single-base tower cost of different types of projects, and standardize the historical data of the single-base tower cost to obtain the original data.

Optionally, the original data is historical basic data related to the engineering cost of different types of transmission lines.

Optionally, the collection of historical data of the single-base tower cost of different types of projects, and the standardized processing of the historical data of the single-base tower cost, including:

Use principal component analysis to reduce the dimension of historical data and output new samples, including: first standardizing the historical data to obtain standardized data; then calculating the correlation coefficient matrix and its eigenvalues and eigenvectors according to the standardized data; finally calculating each For the variance contribution rate of the principal component, the first n principal components whose cumulative variance contribution rate reaches the preset value are selected as the new sample output.

Optionally, building an empirical parameter prediction model in combination with the empirical parameters of the least squares support vector machine model, including:

Given the empirical parameters of the least squares support vector machine model, an empirical parameter prediction model is constructed.

Optionally, the particle swarm optimization algorithm is used to optimize the parameters in the prediction model, and the particle swarm optimization least squares support vector machine prediction model is obtained, including:

First obtain the initial particle speed and position, calculate the current particle fitness according to the particle speed and position, select the position corresponding to the minimum value of the current particle fitness as the particle individual extremum, and compare the fitness of the current particle individual extremum with that of the previous particle individual. The fitness of the extreme value, select the position corresponding to the particle with small fitness as the global extreme value; then update the speed and position of the particle, and repeat the steps of calculating the current particle fitness according to the speed and position of the particle, until the number of iterations When the maximum number of iterations is reached or the precision reaches the preset precision, the global optimal position is output, which is the parameters of the optimized least squares support vector machine; finally, the optimized parameter prediction model is constructed according to the optimized parameters.

Optionally, the original data is used as an input variable to be input into the optimized vector machine prediction model, and the vector machine prediction model is trained to obtain the cost prediction value of a single base tower, including:

A standard LSSVM model is established, and the data is substituted into the model for training. The basic parameter penalty coefficient and kernel function parameters of the _LSSVM prediction model are selected as empirical parameters C=100, σ ² =0.4, and the radial basis ( RadialBasis) kernel function:

Optionally, the machine prediction model performs a training step to obtain a final prediction result.

The technical solutions provided by this application include the following beneficial technical effects:

The present application provides an intelligent prediction method for transmission line engineering cost based on a single base tower. The method includes the following steps: combining the empirical parameters of the least squares support vector machine model to construct an empirical parameter prediction model; The parameters in the prediction model are optimized to obtain a particle swarm optimized least squares support vector machine prediction model; the original data is used as an input variable to be input into the optimized vector machine prediction model, and the vector machine is predicted The model is trained to obtain the cost prediction value of a single base tower. The method provided in this application uses the principal component analysis method to reduce the dimensionality of the indicators, innovatively introduces the particle swarm algorithm to optimize the parameters of the least squares support vector machine model to obtain the optimal parameters, and import the obtained principal component data into the Training and forecasting in the empirical parameter forecasting model and the optimized parameter forecasting model can improve the accuracy of cost forecasting, improve the level of refinement of cost management, and solve the problems of unscientific forecasting methods and low forecasting accuracy.

Description of drawings

In order to illustrate the technical solutions of the present application more clearly, the accompanying drawings that need to be used in the embodiments will be briefly introduced below. Other figures may also be obtained from these figures.

FIG. 1 is a flowchart of a method for intelligently predicting construction cost of a transmission line based on a single base tower provided by an embodiment of the present application.

Detailed ways

In order to enable those skilled in the art to better understand the technical solutions in the application, the technical solutions in the application examples will be described clearly and completely below with reference to the accompanying drawings in the application examples; Examples are only some of the embodiments of the present application, but not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work shall fall within the protection scope of the present application.

Please refer to FIG. 1. FIG. 1 is a flowchart of a method for intelligently predicting construction cost of a power transmission line based on a single base tower provided by an embodiment of the present application. As shown in FIG. 1, a power transmission line based on a single base tower provided by an embodiment of the present application An intelligent prediction method for engineering cost, the method includes the following steps:

S1: Combine the empirical parameters of the least squares support vector machine model to construct an empirical parameter prediction model.

Specifically include the following steps:

用ω和b分别表示回归函数的权向量和偏移量，根据结构风险最小化原则，寻找ω、b使其最小化，即有：(1) First use a nonlinear mapping Map the samples from the original space to the high-dimensional feature space, and convert the nonlinear estimation function into a linear estimation function of the high-dimensional feature space

Use ω and b to represent the weight vector and offset of the regression function, respectively. According to the principle of structural risk minimization, find ω and b to minimize them, that is:

In the formula: ||ω|| ² is used to control the complexity of the model; c is the regularization parameter, which controls the degree of penalty for out-of-error samples; R _emp is the error control function, that is, the ε-insensitive loss function. LSSVM chooses the square of the error ξ _i as the loss function when optimizing the target, so the optimization problem is:

where ξ _i is the relaxation factor. The corresponding Lagrangian method is established to solve this problem, as follows:

where α _i (i=1,2,...,l) is the Lagrange multiplier. According to the KKT optimization conditions, that is, to find the partial derivatives of L with respect to ω, b, ξ _i , and α respectively, and set them equal to 0, we can obtain:

(2) An arbitrary symmetric function that satisfies the Mercer condition is introduced as the kernel function. The parameters of the kernel function determine the complexity of the distribution of samples in the data space, which has a great impact on the performance of the model. α and b are obtained by the least squares method, and finally the decision function of applying LSSVM to regression analysis of nonlinear functions is obtained as:

S2: Use the particle swarm optimization algorithm to optimize the parameters in the prediction model to obtain the particle swarm optimization least squares support vector machine prediction model.

Specifically include the following steps:

(1) In a d-dimensional search space, X={X ₁ ,X ₂ ,L,X _m } is composed of m particles representing possible solutions to the problem, where X _i ={x _i1 ,x _i2 ,L,x _id } represents the position of the ith particle, d is the number of LSSVM parameters, where d=2. Calculate the mean square error generated by each LSSVM on the training set, and construct the following fitness function to calculate the fitness of the individual:

f(x)=MSE(x)

(2) The flying speed of the particle in the d-dimensional space is defined as V _i ={v _i1 ,v _i2 ,L,v _id }, which is represented by P _i ={P _i1 ,P _i2 ,L,P _id } The best position P _best searched by the particle itself (the corresponding fitness is the smallest), and P _g = {P _g1 , P _g2 , L, P _gd } represents the best position G _best of the entire population, then the i-th particle's Velocity and position updates are determined according to the following formulas:

where ω is the inertia weight factor; t is the number of iterations; c ₁ and c ₂ are the acceleration factors, indicating the step size of the particle flying to its optimal position and the overall optimal position; rand() is in the interval [0,1] uniformly distributed random numbers.

(3) When the number of iterations reaches the maximum number of iterations or the precision reaches a preset precision, the iteration loop is exited and the global optimal parameters are returned. The kernel function parameters and penalty coefficients in the LSSVM model can be optimized by using the particle swarm algorithm, avoiding artificial exhaustive trial and error, and obtaining a better fitting effect.

S3: The original data is used as an input variable to be input into the optimized vector machine prediction model, and the vector machine prediction model is trained to obtain the cost prediction value of a single base tower.

Optionally, before the original data is input to the optimized vector machine prediction model as an input variable, the method further includes:

Collect historical data of the single-base tower cost of different types of projects, and standardize the historical data of the single-base tower cost to obtain the original data.

The historical data of the cost of the single base tower is based on the historical data of 20 groups of 110kV transmission line projects actually settled in a certain region of China in 2018 as samples, and the cost data of the single base tower of the project is sorted and decomposed.

Normalize the collected historical data, including the following steps:

(1) Standardize the historical data

The historical data is normalized according to the following formula:

in,

(2) Calculate the sample correlation coefficient matrix

Assuming that the historical data is still represented by X after standardization, the correlation coefficient of the standardized data is:

in,

(3) Calculate the eigenvalues (λ ₁ ,λ ₂ ,L,λ _p ) of the correlation coefficient matrix R and the corresponding eigenvectors a _i =(a _i1 ,a _i2 ,L,a _ip ),i=1,2, L,p

(4) Select important principal components and get the principal component expression

P principal components can be obtained through principal component analysis, but since the variance of each principal component decreases, the amount of information contained in it decreases accordingly, so in actual analysis, it is generally based on the cumulative contribution rate of each principal component (that is, a principal component). The ratio of the variance to the total variance), select the first k principal components, and generally require the cumulative contribution rate of the first k principal components to reach more than 85%.

Further, the original data is historical basic data related to the engineering cost of different types of transmission lines.

Further, the collection of historical data of the single-base tower cost of different types of projects, and the standardized processing of the historical data of the single-base tower cost, including:

Use principal component analysis to reduce the dimension of historical data and output new samples, including: first standardizing the historical data to obtain standardized data; then calculating the correlation coefficient matrix and its eigenvalues and eigenvectors according to the standardized data; finally calculating each For the variance contribution rate of the principal component, the first n principal components whose cumulative variance contribution rate reaches the preset value are selected as the new sample output.

Further, in combination with the empirical parameters of the least squares support vector machine model, an empirical parameter prediction model is constructed, including:

Given the empirical parameters of the least squares support vector machine model, an empirical parameter prediction model is constructed.

Optionally, the particle swarm optimization algorithm is used to optimize the parameters in the prediction model, and the particle swarm optimization least squares support vector machine prediction model is obtained, including:

First obtain the initial particle speed and position, calculate the current particle fitness according to the particle speed and position, select the position corresponding to the minimum value of the current particle fitness as the particle individual extremum, and compare the fitness of the current particle individual extremum with that of the previous particle individual. The fitness of the extreme value, select the position corresponding to the particle with small fitness as the global extreme value; then update the speed and position of the particle, and repeat the steps of calculating the current particle fitness according to the speed and position of the particle, until the number of iterations When the maximum number of iterations is reached or the precision reaches the preset precision, the global optimal position is output, which is the parameters of the optimized least squares support vector machine; finally, the optimized parameter prediction model is constructed according to the optimized parameters.

Optionally, the original data is used as an input variable to be input into the optimized vector machine prediction model, and the vector machine prediction model is trained to obtain the cost prediction value of a single base tower, including:

A standard LSSVM model is established, and the data is substituted into the model for training. The basic parameter penalty coefficient and kernel function parameters of the _LSSVM prediction model are selected as empirical parameters C=100, σ ² =0.4, and the radial basis ( RadialBasis) kernel function:

Further, the machine prediction model performs the training step to obtain the final prediction result.

The method provided in this application uses the principal component analysis method to reduce the dimensionality of the indicators, innovatively introduces the particle swarm algorithm to optimize the parameters of the least squares support vector machine model to obtain the optimal parameters, and import the obtained principal component data into the Training and forecasting in the empirical parameter forecasting model and the optimized parameter forecasting model can improve the accuracy of cost forecasting, improve the level of refinement of cost management, and solve the problems of unscientific forecasting methods and low forecasting accuracy.

It should be noted that relational terms such as "first" and "second" etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any relationship between these entities or operations. any such actual relationship or sequence exists. Moreover, the terms "comprising", "comprising" or any other variation thereof are intended to encompass a non-exclusive inclusion, whereby an article or device comprising a list of elements includes not only those elements, but also other elements not expressly listed, Or also include elements inherent to such a process, method, article or apparatus. Without further limitation, an element qualified by the phrase "comprising a..." does not preclude the presence of additional identical elements in a process, method, article or apparatus that includes the element.

The above descriptions are only specific embodiments of the present application, so that those skilled in the art can understand or implement the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, this application is not intended to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

It should be understood that the present application is not limited to what has been described above and shown in the accompanying drawings and that various modifications and changes may be made without departing from its scope. The scope of the application is limited only by the appended claims.

Claims (8)

1. A power transmission line engineering cost intelligent prediction method based on a single-base tower is characterized by comprising the following steps:

an empirical parameter prediction model is constructed by combining empirical parameters of a least square support vector machine model;

optimizing parameters in the prediction model by utilizing a particle swarm optimization algorithm to obtain a particle swarm optimization least square support vector machine prediction model;

and inputting the original data serving as an input variable into the optimized vector machine prediction model, and training the vector machine prediction model to obtain the cost prediction value of the single-base tower.

2. The intelligent single-base-tower-based power transmission line construction cost prediction method according to claim 1, wherein before the original data is input to the optimized vector machine prediction model as an input variable, the method further comprises:

and collecting historical single-base tower construction cost data of different types of projects, and carrying out standardized processing on the historical single-base tower construction cost data to obtain the original data.

3. The intelligent single-base-tower-based power transmission line construction cost prediction method according to claim 1, wherein the original data is historical basic data related to different types of power transmission line construction costs.

4. The intelligent prediction method for the construction cost of the power transmission line based on the single-base tower as claimed in claim 2, wherein the collecting historical data of the construction cost of the single-base tower of different types of projects and the standardizing the historical data of the construction cost of the single-base tower comprises:

using principal component analysis to reduce dimension of historical data, and outputting a new sample, comprising: firstly, standardizing the historical data to obtain standardized data; then, calculating a correlation coefficient matrix, and a characteristic value and a characteristic vector thereof according to the standardized data; and finally, calculating the variance contribution rate of each principal component, and selecting the first n principal components with the accumulated variance contribution rate reaching a preset value as new samples to be output.

5. The intelligent single-base-tower-based power transmission line engineering cost prediction method according to claim 1, wherein the construction of the empirical parameter prediction model by combining empirical parameters of a least squares support vector machine model comprises:

and (3) giving empirical parameters of the least square support vector machine model, and constructing an empirical parameter prediction model.

6. The intelligent single-base-tower-based power transmission line engineering cost prediction method according to claim 1, wherein the parameters in the prediction model are optimized by using a particle swarm optimization algorithm to obtain a particle swarm optimization least squares support vector machine prediction model, and the method comprises the following steps:

firstly, acquiring the speed and the position of an initialized particle, calculating the fitness of the current particle according to the speed and the position of the particle, selecting the position corresponding to the minimum value of the fitness of the current particle as an individual extreme value of the particle, comparing the fitness of the individual extreme value of the current particle with the fitness of the individual extreme value of the previous particle, and selecting the position corresponding to the particle with low fitness as a global extreme value; then updating the speed and the position of the particle, and repeatedly executing the step of calculating the current particle fitness according to the speed and the position of the particle until the iteration times reach the maximum iteration times or the precision reaches the preset precision, outputting a global optimal position, namely the parameter of the optimized least square support vector machine; and finally, constructing an optimized parameter prediction model according to the optimized parameters.

7. The intelligent single-base-tower-based power transmission line engineering cost prediction method according to claim 1, wherein the step of inputting the original data as input variables into the optimized vector machine prediction model and training the vector machine prediction model to obtain the predicted single-base-tower cost value comprises the steps of:

establishing a standard LSSVM model, substituting data into the model for training, selecting an empirical parameter C of 100, and selecting a basic parameter penalty coefficient and a kernel function parameter of the LSSVM prediction model²0.4 kernel function K (x, x)_i) Choosing a Radial Basis kernel function:

8. the intelligent single-base-tower-based power transmission line engineering cost prediction method according to claim 7, characterized by repeating the steps of inputting the original data as an input variable into the optimized vector machine prediction model, and training the vector machine prediction model to obtain a final prediction result.

Images