WO2023035245A1

WO2023035245A1 - Risk early-warning method applied to electricity market price

Info

Publication number: WO2023035245A1
Application number: PCT/CN2021/117800
Authority: WO
Inventors: 邢劲; 韦仲康; 朱天博; 刘俊; 牛家强; 张斓曦; 马延杰
Original assignee: 冀北电力交易中心有限公司; 四川中电启明星信息技术有限公司
Priority date: 2021-09-10
Filing date: 2021-09-10
Publication date: 2023-03-16

Abstract

The present invention relates to the technical field of electricity information. Disclosed is a risk early-warning method applied to the electricity market price. The method comprises: step A, establishing an electricity transaction price risk factor index model according to a multi-factor noise-reduction weight analysis algorithm; step B, hierarchically exploring an electricity transaction price risk according to the electricity transaction price risk factor index model and a GM (1,1) hybrid variable model, and dividing the electricity transaction price risk into different electricity transaction price risk levels; and step C, establishing an electricity transaction price risk early-warning model according to the electricity transaction price risk factor index model and the electricity transaction price risk levels. By means of the present invention, when an electricity price fluctuation amplitude exceeds a rational threshold value, an electricity market structure can be optimized, and an electricity price transaction behavior of an electricity market can be strengthened.

Description

A risk early warning method applied to electricity market price

technical field

The present invention relates to the field of electric power information technology, specifically, it is a risk warning method applied to electric power market prices, which can optimize the electric power market structure and strengthen the electric price transactions in the electric power market when the electric price fluctuation range exceeds a reasonable threshold Behavior.

Background technique

In the electricity market, all bidding results are ultimately reflected in market prices, and fluctuations in electricity prices directly reflect market risks. Price risk early warning is an important part of electricity market risk management, including two parts: electricity price level forecast and electricity price fluctuation risk measurement. It is necessary for electricity market operators and regulators to make predictions and early warnings for future electricity price levels and their fluctuation risks based on market information. .

On the one hand, in electricity transactions in the electricity market, spot transactions usually exist as a supplementary means for medium and long-term electricity transactions, but they face greater price risks and less credit risks when negotiating between transaction parties. Therefore, there is an urgent need to A risk early warning method applied to electricity market prices to optimize the market structure and strengthen the monitoring, punishment and early warning mechanism of transaction behavior.

On the one hand, in the electricity transaction in the electricity market, the abnormal electricity price fluctuation is an important factor affecting the normal economy of the electricity market. Market members with market power endanger market competition behavior through market power, resulting in abnormally high or low prices, which affect the normal economic and social development order. Therefore, there is an urgent need for a risk warning method applied to electricity market prices. Accelerate the construction of a unified national electricity market, avoid the formation of local market forces by expanding the scope of transactions, and implement scientific demonstrations of the electricity market mechanism and trading rules to achieve stable and efficient operation of the electricity market.

Contents of the invention

The purpose of the present invention is to provide a risk early warning method applied to electricity market prices, realize the monitoring, punishment and early warning functions of electricity price transactions, and have the effect of optimizing the structure of the electricity market and strengthening the electricity price transaction behavior of the electricity market.

The present invention is achieved through the following technical solutions: a risk early warning method applied to electricity market prices, comprising:

Step A. Establish an index model of power trading price risk factors according to the multi-factor noise reduction weight analysis algorithm;

Step B. According to the power transaction price risk factor index model and the GM (1,1) mixed variable model, the power transaction price risk is classified and explored, and divided into different power transaction price risk levels;

Step C. Establish a power transaction price risk early warning model based on the power transaction price risk factor index model and the power transaction price risk level.

In this technical solution, all bidding results in the electricity market will eventually be reflected in market prices, and fluctuations in electricity price levels directly reflect market risks. The price of electricity transactions can be divided into normal price fluctuations and abnormal price fluctuations depending on the cause. Normal transaction price fluctuations are the fluctuations in the prices finally reached in centralized bidding transactions and bilateral negotiation transactions due to changes in the marginal cost of generating units and changes in market supply and demand during the normal operation of the electricity market. For the case of small fluctuations, it is a normal market phenomenon, which can encourage power generation companies, electricity sales companies, and large users participating in centralized bidding transactions and bilateral negotiation transactions to improve their market transaction decision-making capabilities, including improving fuel market prices and market supply and demand levels. , market price prediction accuracy, etc. Abnormal electricity price fluctuation refers to the deviation of electricity price from a reasonable level, which cannot reflect the real market supply and demand relationship and corresponding costs. When the fluctuation of electricity price exceeds a reasonable threshold, the demand for electricity purchase and consumption of electricity sales companies and users will shrink, which will affect the normal order of economic and social development. Relevant market entities are unable to perform the contract, and large users and ordinary users who have given up the right to choose will switch to the grid company for guaranteed power supply at government-controlled prices. At this time, the normal production and operation of grid companies will be seriously affected, which in turn will affect the normal investment of the grid and the power system. orderly operation, etc. Therefore, this technical solution uses the multi-factor noise reduction weight analysis algorithm and the price risk classification technology based on the gray system theory of GM (1,1) mixed variables to build a power transaction price risk early warning system, optimize the market structure, strengthen the monitoring of trading behavior, Punishment and early warning mechanisms, speed up the construction of a unified national electricity market, and avoid the formation of local market forces by expanding the scope of transactions. At the same time, the market mechanism and trading rules are scientifically demonstrated and implemented to realize the stable and efficient operation of the market.

In order to better realize the present invention, further, step A specifically includes:

A1. Determine the influencing factors of power transaction price risk, and list them in a sliced matrix to generate matrix B corresponding to table a1 of the power transaction price risk factor index system;

A2. Self-built index factor analysis system table a2, and further expand the self-built index factor analysis system table a2 to form the required area designated index measurement system table a3;

A3..Based on the multi-factor noise reduction weight analysis algorithm in step A, use the multivariate sorting method and noise reduction verification method to assign values to the power transaction price risk factor index system table a1 in step A1, and combine the index measurement system in step A2 Table a3 generates the power transaction price risk factor index model.

In this technical solution, the multi-factor noise reduction weight analysis technology and the price risk classification technology based on GM(1,1) mixed variables are established. Firstly, the factors affecting the price risk of power trading are expanded in slices to form a risk factor index system. Through the self-built index factor system, including power market structure, market maturity, auxiliary service market, transmission constraints, market rules, load and total installed capacity of the market, and weather conditions, etc., further expand and form a unique index measurement system for the northern Hebei market. And use the multi-factor noise reduction weight analysis technology to assign weight values to the index weights.

In order to better realize the present invention, further, step A3 specifically includes:

Calculate the value of the consistency index CI of matrix B in step A1 according to the multi-factor noise reduction weight analysis algorithm.

In this technical solution, the multi-factor denoising weight analysis technique refers to calculating the weights of the order of importance of factors related to a certain factor in the previous level according to the judgment matrix. Therefore,

In order to better realize the present invention, further, the formula of the multi-factor denoising weight analysis algorithm is: BW=λ _max W, where λ _max is the maximum characteristic root of matrix B, and W is the normalized feature corresponding to λ _max vector;

The formula CI of the consistency index of the matrix B is:

The sum of n eigenvalues of matrix B is equal to n, that is, the consistency index CI of matrix B is equivalent to the average value of the other n-1 eigenvalues except λ _max ;

Verifying the value of the consistency index CI corresponding to the matrix B and adjusting the matrix B according to the value of the consistency index CI.

In this technical solution, when the judgment matrix has complete consistency, CI=0. The larger the λ _max -n, the larger the CI, and the worse the consistency of the judgment matrix. Note that the sum of the n eigenvalues of the matrix B is exactly equal to n, so CI is equivalent to the average value of the other n-1 eigenvalues except λ _max . In order to test whether the judgment matrix has satisfactory consistency, it is necessary to find out the standard for measuring the consistency index CI of matrix B.

In order to better realize the present invention, further, to judge the degree of consistency of the matrix B, when there is complete consistency, the CI is 0, and when the consistency of the judgment matrix B decreases, λ _max -n increases, and the CI is linear thereupon increase;

The standard threshold value of the consistency index CI for measuring the matrix B is preset, and if the threshold value is exceeded, it is judged that the matrix B has reached the standard consistency; otherwise, the matrix B is adjusted.

In this technical solution, the standard threshold value of the consistency index CI of the preset measurement matrix B is RI. For the first-order and second-order judgment matrices, RI is only formal. According to our definition of the judgment matrix, the first-order, second-order The second-order judgment matrix is always completely consistent. When the order is greater than 2, the ratio of the consistency index CI of the judgment matrix to the index RI of the average random consistency of the same order

Called the random consistency ratio of the judgment matrix, denoted as CR. when

When , the judgment matrix has satisfactory consistency, otherwise the judgment matrix needs to be adjusted.

In order to better realize the present invention, further, the multivariate sorting method in step A3 includes:

The default ordering of n elements on the k-1th layer relative to the total target is:

Among them, w ^(k-1) is the component of the normalized feature vector W, that is, the weight of the single sorting of n elements on the k-1th layer;

The single sorting vector of the preset k-th layer n _k elements for the j-th element on the k-1th layer as the criterion is:

Among them, j=1,2...,nk=1,2,..., _nk ;

The weight of elements not dominated by the jth element is zeroed to obtain a matrix of order n _k ×n:

Where the jth column in U ^(k) is the kth layer n _k elements,

For the single sorting vector with the jth element as the criterion on the k-1th layer, the total ranking of each element on the kth layer to the total target is recorded as:

And get the multivariate sorting formula:

Wherein, i=1, 2, . . . , n _k .

In this technical solution, the total order of each element on the kth layer to the total target is:

In order to better realize the present invention, further, the noise reduction verification method in step A3 includes:

Carry out noise reduction verification layer by layer from the top down;

Let some factors in the k-th layer single-order the noise reduction index of the j-th element in the k-1 layer to be

The average random consistency index is:

Then the noise reduction ratio formula for the total ranking of the kth layer is:

In order to better realize the present invention, further, assign values to the power transaction price risk factor index system in step A1 according to the multivariate ranking formula obtained by the multivariate ranking method and the noise reduction ratio formula obtained by the noise reduction verification method.

In order to better realize the present invention, further, step C includes:

According to the power transaction price risk factor index model and the power transaction price risk level, the power transaction price information is collected, and it is judged whether the power transaction price information is the preset warning signal indicator threshold, and if so, the warning signal is output and analyzed.

Compared with the prior art, the present invention has the following advantages and beneficial effects:

(1) To optimize the market structure, strengthen the monitoring, punishment and early warning mechanism of transaction behaviors, accelerate the construction of a unified national electricity market, and avoid the formation of local market power by expanding the scope of transactions;

(2) Carry out scientific demonstration of the market mechanism and trading rules and implement them to realize the stable and efficient operation of the market.

Description of drawings

The present invention will be further described in conjunction with the following drawings and embodiments, and all conceptual innovations of the present invention should be regarded as the disclosed content and the protection scope of the present invention.

FIG. 1 is a flowchart of a risk warning method applied to electricity market prices according to the present invention.

Detailed ways

Example 1:

A risk early warning method applied to electricity market prices in this embodiment, as shown in Figure 1, includes:

In this embodiment, all bidding results in the electricity market are ultimately reflected in market prices, and electricity price level fluctuations directly reflect market risks, and electricity transaction prices can be divided into normal price fluctuations and abnormal price fluctuations depending on the cause. Normal transaction price fluctuations are the fluctuations in the prices finally reached in centralized bidding transactions and bilateral negotiation transactions due to changes in the marginal cost of generating units and changes in market supply and demand during the normal operation of the electricity market. For the case of small fluctuations, it is a normal market phenomenon, which can encourage power generation companies, electricity sales companies, and large users participating in centralized bidding transactions and bilateral negotiation transactions to improve their market transaction decision-making capabilities, including improving fuel market prices and market supply and demand levels. , market price prediction accuracy, etc. Abnormal electricity price fluctuation refers to the deviation of electricity price from a reasonable level, which cannot reflect the real market supply and demand relationship and corresponding costs. When the fluctuation of electricity price exceeds a reasonable threshold, the demand for electricity purchase and consumption of electricity sales companies and users will shrink, which will affect the normal order of economic and social development. Relevant market entities are unable to perform the contract, and large users and ordinary users who have given up the right to choose will switch to the grid company for guaranteed power supply at government-controlled prices. At this time, the normal production and operation of grid companies will be seriously affected, which in turn will affect the normal investment of the grid and the power system. orderly operation, etc. Therefore, this embodiment uses the multi-factor noise reduction weight analysis algorithm and the price risk classification technology based on the gray system theory of GM (1,1) mixed variables to build a power transaction price risk early warning system, optimize the market structure, strengthen the monitoring of trading behavior, Punishment and early warning mechanisms, speed up the construction of a unified national electricity market, and avoid the formation of local market forces by expanding the scope of transactions. At the same time, the market mechanism and trading rules are scientifically demonstrated and implemented to realize the stable and efficient operation of the market.

Example 2:

This embodiment is further optimized on the basis of Embodiment 1, and step A specifically includes:

A1. Determine the influencing factors of electricity transaction price risk, and list them in slices to generate an index system of electricity transaction price risk factors;

A2. Self-built indicator factor analysis system table, and further expand the self-built indicator factor analysis system to form an indicator measurement system designated by the required area;

A3. Based on the multi-factor noise reduction weight analysis algorithm in step A1, use the multivariate sorting method and noise reduction verification method to assign values to the power transaction price risk factor index system, and combine the index factor analysis system table in step A2 to generate index factor weights Analysis system table.

In this embodiment, the power transaction price risk influencing factors are listed in slices, including the list of influencing factor index categories, variable symbols and variable descriptions, including power market structure, market maturity, ancillary service market, transmission constraints, market rules, Load size and climate conditions, among them, the power market structure includes the proportion of thermal power installed capacity and its variable symbol X1, the proportion of hydropower installed capacity and its variable symbol X2, the proportion of clean energy installed capacity and its variable symbol X3 and Electricity coal price and its variable symbol X4, power market maturity includes power generation company HHI and its variable symbol X5, power user HHI and its variable symbol X6, electricity sales company HHI and its variable symbol X7, Top-m share and its variable symbol X8, the year-on-year growth rate of market-oriented trading electricity and its variable symbols X9, the market-oriented participation rate of power generation companies and their variable symbols X10, the market-oriented participation rate of power users and their variable symbols X11, and the market-oriented participation rate of electricity sales companies and their variable symbols X12, the ancillary service market includes whether the ancillary service market is running or not and its variable symbol X13, the transmission constraint includes whether there is a transmission constraint and its variable symbol X14, the market rule includes whether to operate the suppression electricity price policy and its variable symbol X15, and the load size includes the electricity load Size and its variable symbol X16, climate conditions include whether it is a wet season and its variable symbol X17, whether it is a dry season and its variable symbol X18, and whether it is a normal water period and its variable symbol X19.

In this embodiment, the self-built index factor system is constructed, and the self-built index factor system is further expanded to form an index measurement system designated by the required area. Through the self-built index factor system, including power market structure, market maturity, auxiliary The service market, transmission constraints, market rules, load and total installed capacity of the market, and climate conditions are further expanded to form a specific regional market-specific index measurement system, such as the North Hebei index measurement system. Other parts of this embodiment are the same as those of Embodiment 1, so details are not repeated here.

Example 3:

This embodiment is further optimized on the basis of the above-mentioned embodiment 1 or 2. In this embodiment, the multi-factor noise reduction weight analysis algorithm refers to calculating according to the judgment matrix B. The weight of the order of importance of the linked factors. Other parts of this embodiment are the same as those of Embodiment 1 or 2 above, so details are not repeated here.

Example 4:

This embodiment is further optimized on the basis of any one of the above-mentioned embodiments 1-3, and step A3.4 specifically includes:

When the judgment matrix B has complete consistency, CI is 0;

The worse the consistency of the judgment matrix B, the larger the λ _max -n, and the larger the CI;

In this embodiment, the average random consistency index of matrices 1 to 9 is introduced. Among the nine-order indexes, for the first-order and second-order judgment matrices, the average random consistency index RI of the same order is only formal , because the first-order and second-order judgment matrices are always completely consistent. When the order is greater than 2, the ratio of the consistency index CI of the judgment matrix to the index CI of the average random consistency of the same order

Called the random consistency ratio of the judgment matrix, denoted as CR. when

When , the judgment matrix has the consistency up to the standard, that is, when the random consistency ratio of the matrix is less than 0.01, the judgment matrix has the consistency up to the standard, otherwise the judgment matrix needs to be adjusted.

Other parts of this embodiment are the same as those of any one of Embodiments 1-3 above, so details are not repeated here.

Example 5:

This embodiment is further optimized on the basis of the above-mentioned embodiment 4, when the judgment matrix has complete consistency, CI=0. The larger the λ _max -n, the larger the CI, and the worse the consistency of the judgment matrix. Note that the sum of the n eigenvalues of the matrix B is exactly equal to n, so CI is equivalent to the average value of the other n-1 eigenvalues except λ _max . In order to test whether the judgment matrix has satisfactory consistency, it is necessary to find out the standard for measuring the consistency index CI of the matrix B. The standard threshold for the consistency index CI of the measurement matrix B is preset to be RI. For the first-order and second-order judgment matrices , RI is only formal. According to our definition of the judgment matrix, the first-order and second-order judgment matrices are always completely consistent. When the order is greater than 2, the ratio of the consistency index CI of the judgment matrix to the index RI of the average random consistency of the same order

Called the random consistency ratio of the judgment matrix, denoted as CR. when

, the judgment matrix has a satisfactory consistency; otherwise, the judgment matrix needs to be adjusted. The other parts of this embodiment are the same as any of the above-mentioned embodiments 1-4, so details are not repeated here.

Embodiment 6:

This embodiment is further optimized on the basis of the above-mentioned embodiment 2. The multivariate sorting method is to use the results of single sorting of all levels in the same level to calculate the weights of the importance of all factors in this level for the previous level, and It needs to be carried out layer by layer from top to bottom, and other parts of this embodiment are the same as any of the above-mentioned embodiments 1-5, so details are not repeated here.

Embodiment 7:

This embodiment is further optimized on the basis of the above-mentioned embodiment 2. The total order of the elements on the kth layer to the total target is:

The noise reduction check in this embodiment is performed from the top layer down, and the noise reduction check is performed layer by layer. Let some factors in the k-th layer single-order the noise reduction index of the j-th element in the k-1 layer to be

The average random consistency index is

When having nothing to do with the jth element of the k-1 layer in the k layer, this situation need not be considered, so the noise reduction ratio formula of the k layer will be obtained. One item is the same, so it will not be repeated here.

Embodiment 8:

This embodiment is further optimized on the basis of any one of the above-mentioned embodiments 1-7. According to the multiple ranking formula obtained by the multiple ranking method and the noise reduction ratio formula obtained by the noise reduction verification method, the power transaction price risk factor index in step A1 System assignment also uses the reduction method of homogeneous exponential function and some differential and some differential characteristics, estimates parameters based on differential equations, and carries out risk classification based on the time response formula obtained from differential equations. Usually, it first indicates the change range of the power user personal credit risk warning factor index. If the condition of the indicator turns better, there is no need to give an early warning to the indicator; otherwise, it gets worse. Normal transaction prices will not fluctuate too much. Even if the cost of power generation companies is too high due to the sharp rise in fuel prices, resulting in high market transaction prices and large fluctuations in electricity prices, it is a relatively slow formation process. According to the rising or falling trend of the average price in the past 3 months and 12 months, the power transaction price risk level is divided into major risk, relatively large risk, general risk and low risk. If the month is n, the average price of the three months [n-2,n] is higher or lower than the average price of the three months [n-3,n-1] by more than 10% and [n-11,n ] The average price of twelve months is more than 5% higher or lower than the average price of [n-12,n-1] twelve months, then the transaction price is at a major risk level; if [n-2,n ] The average price of the three months is more than 10% higher or lower than the average price of [n-3,n-1], the transaction price is at a relatively high risk level; if [n-11,n] If the average price of twelve months increases or decreases by more than 5% compared with the average price of [n-12,n-1] twelve months, the transaction price is at the general risk level; if [n-2,n] The three-month average price has increased or decreased by less than 10% compared to the [n-3,n-1] three-month average price and the [n-11,n] twelve-month average price has increased or decreased compared to [n -12,n-1] The twelve-month average price increase or decrease is less than 5%, then the transaction price is at a low risk level. In this way, the risk level matching is carried out while assigning the value of the power transaction price risk factor index system. Other parts of this embodiment are the same as any one of the above-mentioned embodiments 1-7, so they are not repeated here.

Embodiment 9:

This embodiment is further optimized on the basis of the above-mentioned embodiment 1. The construction mechanism of the power trading price risk early warning system is mainly to build a system for early warning. Mainly through the index system formed by the multi-factor noise reduction weight analysis technology as the input layer, the index threshold is formed, the early warning signal is output, and the electric power transaction price risk early warning system is constructed. First, starting from index setting and information collection, focusing on constructing early warning models and determining early warning boundaries, and ending with issuing early warning signals and analyzing warning signs, and finally the trading center conducts reasonable regulation and decision-making according to warning signs. Other parts of the embodiment are the same as any one of the above-mentioned embodiments 1-8, so they are not repeated here.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention in any form. Any simple modifications and equivalent changes made to the above embodiments according to the technical essence of the present invention all fall within the scope of the present invention. within the scope of protection.

Claims

A risk early warning method applied to electricity market prices, characterized in that it includes:

Step A. Establish an index model of power trading price risk factors according to the multi-factor noise reduction weight analysis algorithm;

Step B. According to the power transaction price risk factor index model and the GM (1,1) mixed variable model, the power transaction price risk is classified and explored, and divided into different power transaction price risk levels;

Step C. Establish a power transaction price risk early warning model based on the power transaction price risk factor index model and the power transaction price risk level.
A risk warning method applied to electricity market prices according to claim 1, wherein step A specifically includes:

A1. Determine the influencing factors of power transaction price risk, and list them in a sliced matrix to generate matrix B corresponding to table a1 of the power transaction price risk factor index system;

A2. Self-built index factor analysis system table a2, and further expand the self-built index factor analysis system table a2 to form the required area designated index measurement system table a3;

A3. Based on the multi-factor noise reduction weight analysis algorithm in step A, use the multivariate sorting method and noise reduction verification method to assign values to the power transaction price risk factor index system table a1 in step A1, and combine the index measurement system table in step A2 a3 generate the power transaction price risk factor index model.
A risk early warning method applied to electricity market prices according to claim 2, wherein step A3 specifically includes:

Calculate the value of the consistency index CI of matrix B in step A1 according to the multi-factor noise reduction weight analysis algorithm.
A risk early warning method applied to electricity market prices according to claim 3, characterized in that it includes: the multi-factor noise reduction weight analysis algorithm formula is: BW=λ max W, where λ max is the matrix B The largest eigenvalue, W is the normalized eigenvector corresponding to λ max ;

The formula CI of the consistency index of the matrix B is:
The sum of n eigenvalues of matrix B is equal to n, that is, the consistency index CI of matrix B is equivalent to the average value of the other n-1 eigenvalues except λ max ;

Verifying the value of the consistency index CI corresponding to the matrix B and adjusting the matrix B according to the value of the consistency index CI.
A risk early warning method applied to electricity market prices according to claim 4, characterized in that it includes: judging the degree of consistency of matrix B, when there is complete consistency, CI is 0, when judging the consistency of matrix B When the property decreases, the λ max -n increases, and the CI increases linearly;

The standard threshold value of the consistency index CI for measuring the matrix B is preset, and if the threshold value is exceeded, it is judged that the matrix B has reached the standard consistency; otherwise, the matrix B is adjusted.
A risk early warning method applied to electricity market prices according to claim 2, wherein the multivariate sorting method in step A3 includes:

The default ordering of n elements on the k-1th layer relative to the total target is:
Among them, w (k-1) is the component of the normalized feature vector W, that is, the weight of the single sorting of n elements on the k-1th layer;

The pre-set k-th layer n k elements for the j-th element on the k-1th layer as a single sorting vector is
Its expression is:
Among them, j=1,2...,nk=1,2,..., nk ;

The weights of elements not dominated by the jth element are zeroed to obtain n k ×n order matrix U (k) , whose expression is:
Among them, the jth column in the U (k) matrix is the n k element of the kth layer, and for the single sorting vector with the jth element on the k-1th layer as the criterion, the total value of each element on the kth layer to the total target The sort is recorded as:
And get the multivariate sorting formula:
Wherein, i=1, 2, . . . , n k .
A risk early warning method applied to electricity market prices according to claim 2, characterized in that the noise reduction verification method in step A3 includes:

Carry out noise reduction verification layer by layer from the top down;

Let some factors in the k-th layer single-order the noise reduction index of the j-th element in the k-1 layer to be
The average random consistency index is:
Then the noise reduction ratio formula for the total ranking of the kth layer is:
A risk early warning method applied to electricity market prices according to any one of claims 1-7, characterized in that it includes: the multivariate sorting formula obtained by the multivariate sorting method and the noise reduction ratio formula obtained by the noise reduction verification method Assign values to the power transaction price risk factor index system in step A1.
A kind of risk warning method that is applied to electricity market price according to claim 1, is characterized in that, described step C comprises:

According to the power transaction price risk factor index model and the power transaction price risk level, the power transaction price information is collected, and it is judged whether the power transaction price information is the preset warning signal indicator threshold, and if so, the warning signal is output and analyzed.