An energy scheduling control method and system based on big data analysis

Abstract

The present application discloses an energy scheduling and control method and system based on big data analysis, which relates to the technical field of energy control. The method includes: if there are environmental characteristics in different straight sections or curved sections in the historical environmental characteristics that show unstable changes and belong to the relationship of adjacent straight sections or curved sections, or if there are environmental characteristics in different straight sections and curved sections in the historical environmental characteristics that show stable conditions and belong to the relationship of non-adjacent straight sections or curved sections, extracting the correlation coefficients between environmental changes, line curvature condition conversion, and line straight condition conversion that affect energy loss, predicting the energy loss value affected by environmental and line condition conversion during the energy transmission process of the target line to be measured, so as to perform energy scheduling on the original energy allocation amount. The energy scheduling and control method and system based on big data analysis provided by the present invention can improve the rationality of energy scheduling and allocation.

Description

Technical Field

[0001] The present application relates to the field of energy management and control technology, and in particular to an energy scheduling and control method and system based on big data analysis. Background Art

[0002] With the development of the global economy and the acceleration of industrialization, energy, as a key element supporting the operation of modern society, continues to grow in demand. Therefore, it is crucial to make energy dispatching and allocation more rational, reduce and improve the efficiency of energy distribution management and control.

[0003] In traditional technologies, the influencing factors considered in the detection and analysis of energy loss conditions caused by energy transmission during energy scheduling are single, and the correlation and interference effects between the influencing factors under diverse conditions are not considered, and the final detection results are not further verified, which leads to low efficiency of the final energy scheduling and control results and waste of resources.

[0004] Metal material quality supervision and assessment is a key step to ensure material quality and product safety. It involves comprehensive consideration of multiple aspects, including chemical composition, physical properties, mechanical properties, microstructure and processing performance. Strict quality supervision and assessment can not only improve product quality and safety, but also promote technological innovation, enhance market competitiveness, protect public safety and health, and support sustainable development.

[0005] In traditional technology, the inspection and evaluation of metal material quality is often carried out by comprehensively evaluating the data as a whole, without considering the influence of the correlation between multiple areas in different locations and adjacent conditions, so that the final quality evaluation result of the metal material is not accurate enough only in terms of the overall quality. Summary of the invention

[0006] In order to overcome the above-mentioned deficiencies of the prior art, the present application provides an energy scheduling and control method and system based on big data analysis.

[0007] In a first aspect, the present application provides an energy dispatching and control method based on big data analysis, the method comprising:

[0008] Obtain a target line to be tested that requires energy transmission, predict the energy loss value of the target line to be tested under an ideal environment to obtain a first energy consumption prediction value, obtain all historical test lines and historical environmental characteristics, if there are environmental characteristics located in different straight sections or curved sections in the historical environmental characteristics that are unstable and belong to adjacent straight sections or curved sections, extract the correlation coefficient between the environmental change and the energy loss to obtain a first correlation influence coefficient;

[0009] If there are stable environmental features located in different straight sections and curved sections in the historical environmental features, and they belong to the relationship of non-adjacent straight sections or curved sections, the correlation coefficient between the conversion of the line bending condition and the energy loss is extracted to obtain the first second correlation influence coefficient;

[0010] If there are stable environmental features located in different straight sections and curved sections in the historical environmental features, and they belong to the relationship of non-adjacent straight sections or curved sections, the correlation coefficient between the conversion of the line straight condition and the energy loss is extracted to obtain the second second correlation influence coefficient. According to the first correlation influence coefficient, the first second correlation influence coefficient, and the second second correlation influence coefficient, the energy degree value affected by the environment and the line condition conversion during the energy transmission of the target line to be measured is predicted, and the second energy consumption prediction value is output;

[0011] According to the position distribution between the straight sections and the curved sections existing in all historical detection lines, the comprehensive comparison and verification conditions formed by the line combinations belonging to different position distributions of the straight sections and the curved sections are counted;

[0012] One comparison and verification condition that can be used to verify the energy consumption prediction value with the target line to be measured is selected from the comprehensive comparison and verification conditions to verify or adjust the error of the second energy consumption prediction value. According to the first energy consumption prediction value, the first actual energy consumption prediction value and the second actual energy consumption prediction value are obtained. According to the first actual energy consumption prediction value or the second actual energy consumption prediction value, the original energy distribution amount is adjusted for energy scheduling, and the actual energy demand distribution amount is output.

[0013] Preferably, the target line to be measured that needs to carry out energy transmission, the current operation parameter values of the target line to be measured, and the original energy distribution amount preset for the target line to be measured are obtained, and the damage condition of the target line is detected to obtain the target line damage characteristics;

[0014] The theoretical characteristic data of all lines for energy transmission in an ideal environment are obtained, and according to the theoretical characteristic data, a simulation prediction model one is established;

[0015] The target line to be measured, the original energy distribution amount, the target line damage characteristics, and the current operation parameter values are input into the simulation prediction model one for testing to predict the energy loss value of the target line to be measured in an ideal environment to obtain the first energy consumption prediction value;

[0016] All historical detection lines detected in the historical period, the historical operation parameter values of all historical detection lines, and the historical environmental characteristics of various lines in all historical detection lines are obtained. If there are unstable changes in the environmental characteristics located in different straight sections or curved sections in the historical environmental characteristics, and they belong to the relationship of adjacent straight sections or curved sections, they are counted as the first environmental condition characteristics;

[0017] According to the first environmental condition feature, select the first historical line from all historical detection lines, and obtain the first energy loss values respectively belonging to two adjacent straight sections or curved sections in the first historical line;

[0018] According to the first energy loss value and the corresponding environmental change degree value in the first environmental condition feature, extract the correlation coefficient between the environmental change and the energy loss to obtain the first correlation influence coefficient.

[0019] Preferably, if there are stable environmental features located in different straight sections and curved sections in the historical environmental features and they belong to the relationship of non-adjacent straight sections or curved sections, it is counted as the second environmental condition feature;

[0020] According to the second environmental condition feature, select the second historical line from all historical detection lines, and obtain the second energy loss values respectively belonging to two adjacent positions where the front section is a straight section and the rear section is a curved section in the second historical line;

[0021] According to the second energy loss value and the corresponding environmental stability value in the second environmental condition feature, extract the correlation coefficient between the conversion of the line bending condition and the energy loss to obtain the first second correlation influence coefficient.

[0022] Preferably, obtain the third energy loss values respectively belonging to two adjacent positions where the front section is a curved section and the rear section is a straight section in the second historical line;

[0023] According to the third energy loss value and the corresponding environmental stability value in the second environmental condition feature, extract the correlation coefficient between the conversion of the line straight condition and the energy loss to obtain the second second correlation influence coefficient;

[0024] The first environmental condition feature and the second environmental condition feature are combined into the first set of features to be trained;

[0025] The first correlation influence coefficient, the first second correlation influence coefficient, and the second second correlation influence coefficient are combined into the second set of features to be trained;

[0026] Obtain the historical energy distribution of all historical detection lines, and establish the second energy consumption prediction model according to the historical operation parameter values, the historical energy distribution, the first set of features to be trained, and the second set of features to be trained;

[0027] Detect the current environmental features of the target line to be measured, input the current operation parameter values, the current environmental features, the original energy distribution, and the target line to be measured into the second energy consumption prediction model for testing, so as to predict the energy degree value lost by the target line to be measured during the energy transmission process affected by the environment and the line condition conversion, and output the second energy consumption prediction value.

[0028] Preferably, if there are straight sections and curved sections in all historical detection lines, and the straight section is only located at the head position and the remaining part is all curved sections, it is counted as line condition feature one;

[0029] If the straight section is only located at the middle position and the remaining part is all curved sections, it is counted as line condition feature two;

[0030] If the straight section is only located at the end position and the remaining part is all curved sections, it is counted as line condition feature three;

[0031] If the straight section and the curved section are in a cross relationship in terms of position distribution, it is counted as line condition feature four;

[0032] Count the lines that only have straight sections in all historical detection lines, output line condition feature five, and count the lines that only have curved sections in all historical detection lines, output line condition feature six;

[0033] The line condition feature one and the line condition feature six are combined into the comparison and verification condition one, and the line condition feature two and the line condition feature five are combined into the comparison and verification condition two;

[0034] The line condition feature three, the line condition feature five and the line condition feature six are combined into the comparison and verification condition three, and the line condition feature four, the line condition feature five and the line condition feature six are combined into the comparison and verification condition four;

[0035] The comparison and verification condition one, the comparison and verification condition two, the comparison and verification condition three and the comparison and verification condition four are combined into the comprehensive comparison and verification condition.

[0036] Preferably, select a comparison and verification condition that can be used to verify the energy consumption prediction value with the target line to be measured from the comprehensive comparison and verification condition, and output the preprocessing condition;

[0037] Input the lines, current operating parameter values, current environmental characteristics, and original energy distribution amounts in the preprocessing condition into the energy consumption prediction model two for testing, and output the comparison energy consumption prediction value;

[0038] Preset an energy consumption difference interval value. If the difference between the second energy consumption prediction value and the comparison energy consumption prediction value is within the energy consumption difference interval value, integrate the first energy consumption prediction value and the second energy consumption prediction value to obtain the actual energy consumption prediction value one;

[0039] If the difference between the second energy consumption prediction value and the comparison energy consumption prediction value is outside the energy consumption difference interval value, perform error regulation on the second energy consumption prediction value according to the comparison energy consumption prediction value, and output the third energy consumption prediction value;

[0040] Integrate the first energy consumption prediction value and the third energy consumption prediction value to obtain the actual energy consumption prediction value two;

[0041] According to the actual energy consumption prediction value 1 or the actual energy consumption prediction value 2, energy scheduling is performed on the original energy allocation amount to output the actual energy demand allocation amount.

[0042] Second, an energy dispatching and control system based on big data analysis includes:

[0043] The first energy consumption correlation coefficient extraction unit is used to obtain the target line to be tested that needs to be energy transmitted, predict the energy loss value of the target line to be tested under an ideal environment to obtain a first energy consumption prediction value, obtain all historical detection lines and historical environmental characteristics, and if there are environmental characteristics located in different straight sections or curved sections in the historical environmental characteristics that are unstable and belong to adjacent straight sections or curved sections, extract the correlation coefficient between the environmental change and the energy loss to obtain a first correlation influence coefficient;

[0044] The second energy consumption correlation coefficient extraction unit is used to extract the correlation coefficient between the influence of line bending condition conversion on energy loss to obtain the second correlation influence coefficient 1 if there are environmental characteristics located in different straight sections and curved sections in the historical environmental characteristics that are stable and belong to non-adjacent straight sections or curved sections;

[0045] The energy consumption prediction unit is used to extract the correlation coefficient between the energy loss affected by the line straight condition conversion to obtain the second correlation influence coefficient 2 if there are environmental characteristics located in different straight sections and curved sections in the historical environmental characteristics, and the environmental characteristics are stable and belong to non-adjacent straight sections or curved sections. According to the first correlation influence coefficient, the second correlation influence coefficient 1 and the second correlation influence coefficient 2, the energy degree value affected by the environment and line condition conversion during the energy transmission process of the target line to be tested is predicted, and the second energy consumption prediction value is output;

[0046] The verification condition statistics unit is used to calculate the comprehensive verification conditions to be compared for the combinations of the lines belonging to the straight sections and the curved sections under different position distribution conditions according to the position distribution conditions between the straight sections and the curved sections existing in all historical detection lines;

[0047] The energy dispatching unit is used to select a comparison verification condition from the comprehensive comparison verification conditions that can be used to verify the energy consumption prediction value of the target line to be tested, so as to perform error verification or adjustment on the second energy consumption prediction value, and obtain the actual energy consumption prediction value one and the actual energy consumption prediction value two according to the first energy consumption prediction value, and perform energy dispatch on the original energy allocation according to the actual energy consumption prediction value one or the actual energy consumption prediction value two, and output the actual energy demand allocation.

[0048] Compared with the prior art, the present invention has the following characteristics and beneficial effects:

[0049] By predicting the energy loss degree value of the target line to be measured in an ideal environment and differentiating and predicting the energy loss degree value induced by the diverse conditions of the different position distributions between the straight sections and the curved sections existing in the line and whether the environmental characteristics are in a stable condition or an unstable change in a non-ideal environment, and based on the correlation influence and sustainable development characteristics between the data, the correlation influence coefficients between the environmental change influence, the line bending condition conversion influence, and the line straight condition conversion influence and the energy loss degree value of the target line to be measured are extracted, so as to realize the prediction of the energy loss degree value of the target line to be measured in a non-ideal environment. In order to reduce the large error of the prediction result, a further verification process is carried out, that is, by matching the lines with the same distribution condition between the straight section and the curved section from all historical detection lines, and predicting the energy consumption degree value under the same characteristic information as the target line to be measured, and comparing the predicted energy consumption degree value of the target line to be measured, so as to judge whether the predicted energy consumption degree value of the target line to be measured is accurate. If the error is large, error regulation is carried out. Through the above processing method, the large error of the prediction result is greatly reduced, the correlation between the diverse influence factors is fully considered, and the corresponding verification process is carried out, enhancing the rigor of the entire data processing process. Description of the Drawings

[0050] Figure 1 It is a block diagram of the steps of a method for energy dispatch and control based on big data analysis mainly embodied in this embodiment.

[0051] Figure 2 It is a block diagram of the structure of a system for energy dispatch and control based on big data analysis mainly embodied in this embodiment. Detailed Implementation Manner

[0052] The present invention will be further described in detail below in conjunction with the following embodiments.

[0053] Referring to Figure 1 , a method for energy dispatch and control based on big data analysis, the method includes the following steps:

[0054] S1. Obtain the target line to be measured that needs to carry out energy transmission, predict the energy loss value of the target line to be measured in an ideal environment to obtain the first energy consumption prediction value, obtain all historical detection lines and historical environmental characteristics. If there are environmental characteristics in the historical environmental characteristics that are in an unstable change situation in different straight sections or curved sections and belong to the adjacent straight section or curved section relationship, extract the correlation coefficient between the environmental change and the energy loss to obtain the first correlation influence coefficient.

[0055] S2. If there are environmental features in the historical environmental features that are in a stable state in different straight-line sections and curved sections, and belong to non-adjacent straight-line sections or curved-section relationships, extract the correlation coefficient between the conversion of the line bending condition and the energy loss to obtain the first second correlation influence coefficient.

[0056] S3. If there are environmental features in the historical environmental features that are in a stable state in different straight-line sections and curved sections, and belong to non-adjacent straight-line sections or curved-section relationships, extract the correlation coefficient between the conversion of the line straight condition and the energy loss to obtain the second second correlation influence coefficient. According to the first correlation influence coefficient, the first second correlation influence coefficient, and the second second correlation influence coefficient, predict the energy level value affected by the environment and the line condition conversion during the energy transmission of the target line to be measured, and output the second energy consumption prediction value.

[0057] S4. According to the position distribution between the straight-line sections and the curved sections in all historical detection lines, count the comprehensive comparison and verification conditions formed by the line combinations belonging to different position distributions of the straight-line sections and the curved sections.

[0058] S5. Select a comparison and verification condition that can be used to verify the energy consumption prediction value from the comprehensive comparison and verification conditions to verify or adjust the error of the second energy consumption prediction value. According to the first energy consumption prediction value, obtain the first actual energy consumption prediction value and the second actual energy consumption prediction value. According to the first actual energy consumption prediction value or the second actual energy consumption prediction value, perform energy scheduling on the original energy distribution amount, and output the actual energy demand distribution amount.

[0059] Specifically, by predicting the energy loss degree value of the target line to be measured in an ideal environment and distinguishing and predicting the energy loss degree values induced by the different position distributions between the straight sections and the curved sections existing in the line and the diversity conditions where the environmental characteristics are in a stable condition or an unstable change in a non-ideal environment, and based on the correlation influence and sustainable development characteristics between the data, the correlation influence coefficients between the environmental change influence, the line bending condition conversion influence, and the line straight condition conversion influence and the energy loss degree value of the target line to be measured are extracted to achieve the prediction of the energy loss degree value of the target line to be measured in a non-ideal environment. In order to reduce the large error of the prediction result, a further verification process is carried out, that is, by matching the lines with the same distribution condition between the straight sections and the curved sections from all the historical detection lines, and predicting the energy consumption degree value under the same characteristic information as the target line to be measured, and comparing it with the energy consumption degree value predicted by the target line to be measured to determine whether the predicted energy consumption degree value of the target line to be measured is accurate. If the error is large, error regulation is carried out. Through the above processing method, the large error of the prediction result is greatly reduced, the correlation between the diversity influence factors is fully considered, and the corresponding verification process is carried out to enhance the rigor of the entire data processing process.

[0060] Specifically, step S1 includes the following sub-steps:

[0061] Obtain the target line to be measured for energy transmission, the current operating parameter values of the target line to be measured, and the original energy distribution amount preset for the target line to be measured, and detect the damage condition of the target line to obtain the target line damage characteristics.

[0062] Obtain the theoretical characteristic data of all the lines for energy transmission in an ideal environment, and establish a simulation prediction model I according to the theoretical characteristic data.

[0063] Input the target line to be measured, the original energy distribution amount, the target line damage characteristics, and the current operating parameter values into the simulation prediction model I for testing to predict the energy loss value of the target line to be measured in an ideal environment to obtain the first energy consumption prediction value.

[0064] Obtain all the historical detection lines detected in the historical period, the historical operating parameter values of all the historical detection lines, and the historical environmental characteristics of various lines in all the historical detection lines. If there are unstable changes in the environmental characteristics located in different straight sections or curved sections in the historical environmental characteristics and they belong to the relationship of adjacent straight sections or curved sections, they are counted as environmental condition characteristics I.

[0065] According to the first environmental condition feature, select the first historical line from all historical detection lines, and obtain the first energy loss values respectively belonging to two adjacent straight sections or curved sections in the first historical line.

[0066] According to the first energy loss value and the corresponding environmental change degree value in the first environmental condition feature, extract the correlation coefficient between the environmental change and the energy loss to obtain the first correlation influence coefficient.

[0067] Specifically, such as the current operating parameter values (parameters such as the power of the line operation, capacitance value, resistance, voltage, etc.), Simulation Prediction Model 1 (such as according to all the line characteristics including energy loss values, line lengths, line width radii, line materials, comprehensive line damage degrees, and line transmission operation parameter values, etc. of the lines that have undergone energy transmission detection under ideal conditions (referring to the situation without considering the influence of temperature and humidity in the environment). For example, using the artificial intelligence model in the existing technology: using machine learning algorithms such as neural networks and support vector machines to learn and train a large amount of historical data to establish a mapping relationship between the input (such as the electrical parameters, operating status, environmental factors, etc. of the line) and the output (line loss value). The trained model can predict the corresponding line loss value according to the new input data.For example, a neural network model can automatically learn complex non-linear relationships in data and has good adaptability to changing power operation conditions. The first energy consumption prediction value (if it is Z1), all historical detection lines (if there are A, B, C, D, E, F), the first environmental condition feature (taking A as an example, the historical environmental feature here refers to the temperature and humidity in the external environment. If the temperature is used as an example here, if there are straight sections in A and the environmental temperature values of the straight sections are unstable, such as sub-sections a1, a2, a3 in the straight section, where a1 and a2 are adjacent, a2 and a3 are adjacent, the temperature at a1 is w1, the temperature at a2 is w2, and the temperature at a3 is w3, and the adjacent values among w1, w2, w3 are different. Or if there are curved sections in A, the explanation is the same as that for the case of straight sections. Here, the straight section or the curved section means that there is one of the straight section and the curved section, not both at the same time), the first historical line (if A is selected because there is the first environmental condition feature in A), and obtain the first energy loss values of two adjacent positions in the first historical line that are both straight sections or curved sections respectively (such as obtaining the energy loss values of sections a1 and a2 as s1 and s2 respectively, and a1 and a2 are both straight sections). The first correlation influence coefficient (such as (w2 - w1):(s2 - s1) if it is r0. To improve the accuracy of the data, statistics can be carried out in multiple different historical periods and averaged to obtain the average coefficient R1. It should be noted that the averaging condition here is the case of the same unstable changing environmental temperature value in different historical periods. For the unstable changing environmental temperature and humidity values of different straight sections or curved sections in A, the corresponding correlation influence coefficients obtained are also different. For example, for a1, a2, a3 in A, a1 and a2 are in an unstable changing environmental temperature and humidity situation, and a2 and a3 are in an unstable changing environmental temperature and humidity situation. The correlation influence coefficient obtained in the case of a1 and a2 is one value, and the correlation influence coefficient obtained in the case of a2 and a3 is one value. Finally, a set of the second correlation influence coefficient is obtained, that is, the second correlation influence coefficient includes multiple data sets obtained by statistics, that is, the first correlation influence coefficient, and so on in the following text).

[0068] Specifically, step S2 includes the following sub-steps:

[0069] If there are stable environmental features in different straight sections and curved sections in the historical environmental features and they belong to the relationship of non-adjacent straight sections or curved sections, they are counted as the second environmental condition feature.

[0070] According to the second environmental condition feature, select the second historical line from all historical detection lines and obtain the second energy loss values of two adjacent positions in the second historical line where the front section is a straight section and the rear section is a curved section respectively.

[0071] According to the energy loss value 2 and the corresponding environmental stability value in the environmental characteristic 2, the correlation coefficient between the influence of the line bending condition conversion on the energy loss is extracted to obtain the second correlation influence coefficient 1.

[0072] Specifically, such as environmental characteristic 2 (taking B as an example, if there is a straight line segment in B, and the ambient temperature value of the straight line segment is stable, such as the straight line segments are b1 and b3, and the curved segment is b2, where b1 and b2 are adjacent, b2 and b3 are adjacent, and the temperature of b1, b2, and b3 is w4, this distribution is the environmental characteristic 2), historical route 2 (if B is selected, because there is environmental characteristic 2 in B), energy loss value 2 (such as obtaining the energy loss values ​​of the respective segments of b1 and b2 if they are s3 and s4 respectively), the second correlation influence coefficient 1 (such as statistically calculating the curvature q1 of b2, if q1:(s4-s3) is r1, in order to improve the accuracy of the data, statistics of multiple different historical periods can be performed, and the average value can be calculated to obtain the average coefficient R2, that is, the second correlation influence coefficient 1).

[0073] The specific step S3 includes the following sub-steps:

[0074] Obtain energy loss value three of two adjacent sections in historical route two, with the front section being a curved section and the rear section being a straight section.

[0075] According to the energy loss value three and the corresponding environmental stability value in the environmental characteristic two, the correlation coefficient between the influence of the line straight line condition conversion on the energy loss is extracted to obtain the second correlation influence coefficient two.

[0076] Environmental feature 1 and environmental feature 2 are combined into feature set 1 to be trained.

[0077] The first correlation influence coefficient, the second correlation influence coefficient one and the second correlation influence coefficient two are combined into a second feature set to be trained.

[0078] The historical energy distribution of all historical detection lines is obtained, and the energy consumption prediction model 2 is established according to the historical operation parameter values, the historical energy distribution, the first feature set to be trained and the second feature set to be trained.

[0079] Detect the current environmental characteristics of the target line to be tested, input the current operating parameter values, current environmental characteristics, original energy allocation and the target line to be tested into the second energy consumption prediction model for testing, so as to predict the energy loss value of the target line to be tested due to the change of environment and line conditions during energy transmission, and output the second energy consumption prediction value.

[0080] Specifically, for example, the energy loss value three (such as obtaining the energy loss values of the respective sections of b2 and b3, which are s4 and s5 respectively), the second correlation influence coefficient two (such as q1:(s5 - s4) being r2. To improve the accuracy of the data, statistics can be carried out for multiple different historical periods and averaged to obtain the average coefficient R3, that is, the second correlation influence coefficient two), the energy consumption prediction model two (such as H = KX + Y + W0, where H refers to the energy loss degree value, K refers to the historical energy distribution, Y refers to the energy consumption quota value corresponding to the historical operation parameter value, W0 refers to the energy consumption compensation value corresponding to the environmental temperature and humidity, X: which one to select from the second training feature set, determined by the first training feature set, that is, the first training feature set is used to match the distribution between the straight section and the curved section in the line to correspondingly select the corresponding correlation influence coefficient in the second training feature set, and it should be noted that: X is the sum of the correlation influence coefficients correspondingly matched from the second training feature set. For example, among b1, b2, b3, b4 in B, if the distribution characteristics of b1 and b2 belong to the loop condition feature one, the selected correlation influence coefficient is R1. If the distribution characteristics of b2 and b3 belong to the loop condition feature two, where b2 is the straight section and b3 is the curved section, the selected correlation influence coefficient is R2. If the distribution characteristics of b3 and b4 belong to the loop condition feature two, where b3 is the curved section and b4 is the straight section, the selected correlation influence coefficient is R3, then X is R1 + R2 + R3), the second energy consumption prediction value (match the feature information according to the current operation parameter value, the current environmental characteristics and the target line to be measured with the historical operation parameter value and the first training feature set in the energy consumption prediction model two to determine whether the distribution feature of the target line to be measured belongs to the loop condition feature one or the loop condition feature two, so as to correspondingly match the correlation influence coefficient from the second training feature set and substitute it into H = KX + Y + W0 for testing. If the predicted second energy consumption prediction value is Z2).

[0081] Specific step S4 includes the following sub - steps:

[0082] If there are straight sections and curved sections in all historical detection lines, and the straight section is only located at the head position and the remaining part is all curved sections, it is counted as line condition feature one.

[0083] If the straight section is only located at the middle position and the remaining part is all curved sections, it is counted as line condition feature two.

[0084] If the straight section is only located at the end position and the remaining part is all curved sections, it is counted as line condition feature three.

[0085] If the straight section and the curved section are in a cross - relationship of position distribution, it is counted as line condition feature four;

[0086] Count the lines that only contain straight sections among all historical detection lines, output line condition feature five, and count the lines that only contain curved sections among all historical detection lines, output line condition feature six.

[0087] Line condition feature one and line condition feature six are combined into comparison verification condition one, and line condition feature two and line condition feature five are combined into comparison verification condition two.

[0088] Line condition feature three, line condition feature five and line condition feature six are combined into comparison verification condition three, and line condition feature four, line condition feature five and line condition feature six are combined into comparison verification condition four.

[0089] Comparison verification condition one, comparison verification condition two, comparison verification condition three and comparison verification condition four are combined into a comprehensive comparison verification condition.

[0090] Specifically, for line condition feature one (if there are straight sections and curved sections in A, which are a1 and a2 respectively, where a1 only exists at the head position of A, and the remaining part is a2 curved section, this distribution is line condition feature one), line condition feature two (if there are sub-sections b1, b2, b3 in B, where b1 and b3 are straight sections, b2 is a curved section, and b2 is located at the middle position of B, this distribution is line condition feature two), line condition feature three (if there are sub-sections c1, c2 in C, where c1 is a curved section, c2 is a straight section, and c2 only exists at the head position of C, and the remaining part is c1 curved section, this distribution is line condition feature three), line condition feature four (if there are sub-sections d1, d2, d3, d4, d5 in D, where d1, d3, d5 are all curved sections, d2, d4 are all straight sections, and d1, d2, d3, d4, d5 are adjacent to each other in turn, this distribution is line condition feature four), line condition feature five (such as E is a straight line and there is no curved section), line condition feature six (such as F is a curved line and there is no straight section), comparison verification condition one (such as the combination of A and F is a condition information for subsequent feature matching with the target line to be measured), comparison verification condition two (such as the combination of B and E is a condition information for subsequent feature matching with the target line to be measured), comparison verification condition three (such as the combination of C, E and F is a condition information for subsequent feature matching with the target line to be measured), comparison verification condition four (such as the combination of D, E and F is a condition information for subsequent feature matching with the target line to be measured).

[0091] Specific step S5 includes the following sub-steps:

[0092] Select a comparison verification condition that can be used to verify the energy consumption prediction value with the target line to be measured from the comprehensive comparison verification condition, and output the preprocessing condition.

[0093] Input the circuit, current operating parameter values, current environmental characteristics, and original energy distribution amount in the preprocessing conditions into the second energy consumption prediction model for testing, and output the energy consumption prediction value to be compared.

[0094] Preset the energy consumption difference range value. If the difference between the second energy consumption prediction value and the energy consumption prediction value to be compared is within the energy consumption difference range value, then integrate the first energy consumption prediction value and the second energy consumption prediction value to obtain the first actual energy consumption prediction value.

[0095] If the difference between the second energy consumption prediction value and the energy consumption prediction value to be compared is outside the energy consumption difference range value, then perform error regulation on the second energy consumption prediction value according to the energy consumption prediction value to be compared, and output the third energy consumption prediction value.

[0096] Integrate the first energy consumption prediction value and the third energy consumption prediction value to obtain the second actual energy consumption prediction value.

[0097] According to the first actual energy consumption prediction value or the second actual energy consumption prediction value, perform energy scheduling on the original energy distribution amount, and output the actual energy demand distribution amount.

[0098] Specifically, for example, in the preprocessing conditions (if the target line to be tested is L, and if the sub-sections in L include l1, l2, l3, l4, where l1 and l4 are straight sections, l2 and l3 are curved sections, and the environmental characteristics to which l1, l2, l3, and l4 belong show unstable changes, such as w5, w6, w7, w6 respectively, then this distribution situation conforms to C in the third comparison verification condition, so select the third comparison verification condition as the condition information for subsequent verification and comparison processing, that is, the third comparison verification condition is the preprocessing condition), the energy consumption prediction value to be compared (for example, if F in the third comparison verification condition is used as an example, then F is correspondingly divided into sub-sections f1, f2, f3, f4. According to the current operating parameter values, current environmental characteristics, and F, match them with the first environmental condition characteristics, then select the associated influence coefficients from the first training feature set, and sum the selected multiple associated influence coefficients, and finally substitute them into H = KX + Y + W0 for testing. If the predicted energy consumption prediction value to be compared is Z3), the preset energy consumption difference range value (if it is T1 - T2, including T1 and T2, which is preset based on the energy consumption error values detected and statistically analyzed in the historical period and can be updated in real time), the first actual energy consumption prediction value (if the difference between Z3 and Z2 is between T1 - T2, then Z1 + Z2 is the first actual energy consumption prediction value), the third energy consumption prediction value (if the difference between Z3 and Z2 is outside T1 - T2, that is, greater than T1 - T2, it means that the prediction error is large and can be fine-tuned. For example, if Z2 + (Z2 + Z3) / 2 is Z4, then Z1 + Z4 is the third energy consumption prediction value), the actual energy demand distribution amount (that is, the original energy distribution amount + Z1 + Z4).

[0099] An energy dispatching and control system based on big data analysis, by applying the above-mentioned energy dispatching and control method based on big data analysis, includes an energy consumption correlation coefficient extraction unit 1, an energy consumption correlation coefficient extraction unit 2, an energy consumption prediction unit, a verification condition statistics unit and an energy dispatching unit, with reference to Figure 2 , obtain the target line to be tested that needs to be energy transmitted through the energy consumption correlation coefficient extraction unit 1, predict the energy loss value of the target line to be tested under an ideal environment to obtain a first energy consumption prediction value, obtain all historical detection lines and historical environmental characteristics, if there are environmental characteristics located in different straight sections or curved sections in the historical environmental characteristics, which are unstable and belong to adjacent straight sections or curved sections, extract the correlation coefficient between the impact of environmental changes on energy loss to obtain a first correlation influence coefficient; through the energy consumption correlation coefficient extraction unit 2, if there are environmental characteristics located in different straight sections and curved sections in the historical environmental characteristics, which are stable and belong to non-adjacent straight sections or curved sections, extract the correlation coefficient between the impact of line bending condition conversion on energy loss to obtain a second correlation influence coefficient 1; through the energy consumption prediction unit, if there are environmental characteristics located in different straight sections and curved sections in the historical environmental characteristics, which are stable and belong to non-adjacent straight sections or curved sections, extract the line straight The correlation coefficient between the energy loss affected by the condition conversion is used to obtain the second correlation influence coefficient 2. According to the first correlation influence coefficient, the second correlation influence coefficient 1 and the second correlation influence coefficient 2, the energy degree value affected by the environment and line condition conversion during the energy transmission of the target line to be tested is predicted, and the second energy consumption prediction value is output; the verification condition statistical unit is used to count the comprehensive comparison verification conditions composed of the lines under different position distribution conditions of the straight sections and the curved sections according to the position distribution between the straight sections and the curved sections in all historical detection lines; the energy scheduling unit is used to screen out a comparison verification condition that can be used to verify the energy consumption prediction value of the target line to be tested from the comprehensive comparison verification conditions, so as to perform error verification or adjustment on the second energy consumption prediction value, and according to the first energy consumption prediction value, obtain the actual energy consumption prediction value 1 and the actual energy consumption prediction value 2, and according to the actual energy consumption prediction value 1 or the actual energy consumption prediction value 2, perform energy scheduling on the original energy allocation, and output the actual energy demand allocation.

[0100] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Therefore, any equivalent changes made according to the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. An energy scheduling and control method based on big data analysis, characterized in that, The steps include: Obtain the target line to be measured for energy transmission, predict the energy loss value of the target line to be measured in an ideal environment to obtain the first energy consumption prediction value, obtain all historical detection lines and historical environmental characteristics. If there are unstable changes in the environmental characteristics located in different straight sections or curved sections in the historical environmental characteristics, and they belong to the relationship of adjacent straight sections or curved sections, extract the correlation coefficient between the environmental changes affecting energy loss to obtain the first correlation influence coefficient; If there are stable environmental characteristics located in different straight sections and curved sections in the historical environmental characteristics, and they belong to the relationship of non - adjacent straight sections or curved sections, extract the correlation coefficient between the conversion of the line bending condition affecting energy loss to obtain the first second - correlation influence coefficient; If there are stable environmental characteristics located in different straight sections and curved sections in the historical environmental characteristics, and they belong to the relationship of non - adjacent straight sections or curved sections, extract the correlation coefficient between the conversion of the line straight condition affecting energy loss to obtain the second second - correlation influence coefficient. According to the first correlation influence coefficient, the first second - correlation influence coefficient, and the second second - correlation influence coefficient, predict the degree of energy loss affected by the environment and the conversion of the line condition during the energy transmission of the target line to be measured, and output the second energy consumption prediction value; According to the position distribution of the straight sections and curved sections in all historical detection lines, count the comprehensive comparison and verification conditions formed by the line combinations belonging to different position distributions of the straight sections and curved sections; Select a comparison and verification condition that can be used to verify the energy consumption prediction value from the comprehensive comparison and verification conditions to verify or adjust the error of the second energy consumption prediction value. According to the first energy consumption prediction value, obtain the first actual energy consumption prediction value and the second actual energy consumption prediction value. According to the first actual energy consumption prediction value or the second actual energy consumption prediction value, perform energy scheduling on the original energy allocation amount, and output the actual energy demand allocation amount.

2. The energy scheduling and control method based on big data analysis according to claim 1, wherein, The step of obtaining the target line to be measured for energy transmission, predicting the energy loss value of the target line to be measured in an ideal environment to obtain the first energy consumption prediction value, obtaining all historical detection lines and historical environmental characteristics. If there are unstable changes in the environmental characteristics located in different straight sections or curved sections in the historical environmental characteristics, and they belong to the relationship of adjacent straight sections or curved sections, extracting the correlation coefficient between the environmental changes affecting energy loss to obtain the first correlation influence coefficient is specifically as follows: Obtain the target line to be measured for energy transmission, the current operation parameter values of the target line to be measured, and the original energy allocation amount preset for the target line to be measured, and detect the damage condition of the target line to obtain the target line damage characteristics; Obtain the theoretical characteristic data of all lines for energy transmission in an ideal environment, and establish a simulation prediction model one according to the theoretical characteristic data; Input the target line to be measured, the original energy allocation amount, the target line damage characteristics, and the current operation parameter values into the simulation prediction model one for testing to predict the energy loss value of the target line to be measured in an ideal environment to obtain the first energy consumption prediction value; Obtain all historical detection lines detected in historical periods, the historical operation parameter values of all historical detection lines, and the historical environmental characteristics to which various lines in all historical detection lines belong. If there are unstable changes in the environmental characteristics located in different straight sections or curved sections in the historical environmental characteristics, and they belong to the relationship of adjacent straight sections or curved sections, it is counted as environmental condition feature one; According to the environmental condition feature one, select historical line one from all historical detection lines, and obtain the energy loss values one belonging to two adjacent positions that are both straight sections or curved sections in historical line one; According to the energy loss value one and the environmental change degree value corresponding to the environmental condition feature one, extract the correlation coefficient between the environmental change and the energy loss to obtain the first correlation influence coefficient.

3. The energy scheduling and control method based on big data analysis according to claim 2, characterized in that, If there are stable environmental characteristics located in different straight sections or curved sections in the historical environmental characteristics, and they belong to the relationship of non-adjacent straight sections or curved sections, the step of extracting the correlation coefficient between the conversion of the line bending condition and the energy loss to obtain the second correlation influence coefficient one is as follows: If there are stable environmental characteristics located in different straight sections and curved sections in the historical environmental characteristics, and they belong to the relationship of non-adjacent straight sections or curved sections, it is counted as environmental condition feature two; According to the environmental condition feature two, select historical line two from all historical detection lines, and obtain the energy loss values two belonging to two adjacent positions where the front section is a straight section and the rear section is a curved section in historical line two; According to the energy loss value two and the environmental stability value corresponding to the environmental condition feature two, extract the correlation coefficient between the conversion of the line bending condition and the energy loss to obtain the second correlation influence coefficient one.

4. The energy scheduling and control method based on big data analysis according to claim 3, wherein, If there are stable environmental characteristics located in different straight sections and curved sections in the historical environmental characteristics, and they belong to the relationship of non-adjacent straight sections or curved sections, the step of extracting the correlation coefficient between the conversion of the line straight condition and the energy loss to obtain the second correlation influence coefficient two, and predicting the energy degree value lost by the target line to be measured during the energy transmission process affected by the environment and the conversion of the line condition according to the first correlation influence coefficient, the second correlation influence coefficient one, and the second correlation influence coefficient two, and outputting the second energy consumption prediction value is as follows: Obtain the energy loss values three belonging to two adjacent positions where the front section is a curved section and the rear section is a straight section in historical line two; According to the energy loss value three and the environmental stability value corresponding to the environmental condition feature two, extract the correlation coefficient between the conversion of the line straight condition and the energy loss to obtain the second correlation influence coefficient two; The environmental condition feature one and the environmental condition feature two are combined into the training feature set one; The first correlation influence coefficient, the second correlation influence coefficient one, and the second correlation influence coefficient two are combined into the training feature set two; Obtain the historical energy distribution amounts of all historical detection lines, and establish an energy consumption prediction model two according to the historical operation parameter values, the historical energy distribution amounts, the training feature set one, and the training feature set two; Detect the current environmental characteristics to which the target line to be measured belongs, input the current operating parameter value, current environmental characteristics, original energy distribution quantity, and the target line to be measured into the second energy consumption prediction model for testing, so as to predict the degree value of the energy loss affected by the environmental and line condition conversion during the energy transmission process of the target line to be measured, and output the second energy consumption prediction value.

5. The energy scheduling and control method based on big data analysis according to claim 4, characterized in that, The steps of statistically obtaining the comprehensive comparison and verification conditions formed by the lines combined with different position distributions of the straight sections and curved sections according to the position distribution between the straight sections and curved sections existing in all historical detected lines are as follows: If there are straight sections and curved sections in all historical detected lines, and the straight section is only located at the head end position and the remaining part is all curved sections, it is statistically recorded as line condition feature one; If the straight section is only located at the middle position and the remaining part is all curved sections, it is statistically recorded as line condition feature two; If the straight section is only located at the tail end position and the remaining part is all curved sections, it is statistically recorded as line condition feature three; If the straight section and the curved section are in a position distribution cross relationship, it is statistically recorded as line condition feature four; Statistically obtain the lines that only have straight sections in all historical detected lines, output line condition feature five, and statistically obtain the lines that only have curved sections in all historical detected lines, output line condition feature six; The line condition feature one and the line condition feature six are combined into comparison and verification condition one, and the line condition feature two and the line condition feature five are combined into comparison and verification condition two; The line condition feature three, the line condition feature five, and the line condition feature six are combined into comparison and verification condition three, and the line condition feature four, the line condition feature five, and the line condition feature six are combined into comparison and verification condition four; The comparison and verification condition one, the comparison and verification condition two, the comparison and verification condition three, and the comparison and verification condition four are combined into comprehensive comparison and verification conditions.

6. The energy scheduling and control method based on big data analysis according to claim 5, characterized in that Select one comparison and verification condition that can be used to verify the energy consumption prediction value with the target line to be measured from the comprehensive comparison and verification conditions to perform error verification or adjustment on the second energy consumption prediction value, and obtain the actual energy consumption prediction value one and the actual energy consumption prediction value two according to the first energy consumption prediction value. According to the actual energy consumption prediction value one or the actual energy consumption prediction value two, perform energy scheduling on the original energy distribution quantity and output the actual energy demand distribution quantity. The specific steps are as follows: Select one comparison and verification condition that can be used to verify the energy consumption prediction value with the target line to be measured from the comprehensive comparison and verification conditions, and output the preprocessing condition; Input the line, current operating parameter value, current environmental characteristics, and original energy distribution quantity in the preprocessing condition into the second energy consumption prediction model for testing, and output the comparison energy consumption prediction value; Preset an energy consumption difference range value. If the difference between the second energy consumption prediction value and the comparison energy consumption prediction value is within the energy consumption difference range value, integrate the first energy consumption prediction value and the second energy consumption prediction value to obtain the actual energy consumption prediction value one; If the difference between the second energy consumption prediction value and the comparison energy consumption prediction value is outside the energy consumption difference range value, perform error regulation on the second energy consumption prediction value according to the comparison energy consumption prediction value, and output the third energy consumption prediction value; Integrate the first energy consumption prediction value and the third energy consumption prediction value to obtain an actual energy consumption prediction value 2; According to the actual energy consumption prediction value 1 or the actual energy consumption prediction value 2, energy scheduling is performed on the original energy allocation amount to output the actual energy demand allocation amount.

7. An energy scheduling and control system based on big data analysis, characterized in that, The system is used to implement an energy dispatching and control method based on big data analysis as described in any one of claims 1 to 6, comprising: The first energy consumption correlation coefficient extraction unit is used to obtain the target line to be tested that needs to be energy transmitted, predict the energy loss value of the target line to be tested under an ideal environment to obtain a first energy consumption prediction value, obtain all historical detection lines and historical environmental characteristics, and if there are environmental characteristics located in different straight sections or curved sections in the historical environmental characteristics that are unstable and belong to adjacent straight sections or curved sections, extract the correlation coefficient between the environmental change and the energy loss to obtain a first correlation influence coefficient; The second energy consumption correlation coefficient extraction unit is used to extract the correlation coefficient between the influence of line bending condition conversion on energy loss to obtain the second correlation influence coefficient 1 if there are environmental characteristics located in different straight sections and curved sections in the historical environmental characteristics that are stable and belong to non-adjacent straight sections or curved sections; The energy consumption prediction unit is used to extract the correlation coefficient between the energy loss affected by the line straight condition conversion to obtain the second correlation influence coefficient 2 if there are environmental characteristics located in different straight sections and curved sections in the historical environmental characteristics, and the environmental characteristics are stable and belong to non-adjacent straight sections or curved sections. According to the first correlation influence coefficient, the second correlation influence coefficient 1 and the second correlation influence coefficient 2, the energy degree value affected by the environment and line condition conversion during the energy transmission process of the target line to be tested is predicted, and the second energy consumption prediction value is output; The verification condition statistics unit is used to calculate the comprehensive verification conditions to be compared for the combinations of the lines belonging to the straight sections and the curved sections under different position distribution conditions according to the position distribution conditions between the straight sections and the curved sections existing in all historical detection lines; The energy dispatching unit is used to select a comparison verification condition from the comprehensive comparison verification conditions that can be used to verify the energy consumption prediction value of the target line to be tested, so as to perform error verification or adjustment on the second energy consumption prediction value, and obtain the actual energy consumption prediction value one and the actual energy consumption prediction value two according to the first energy consumption prediction value, and perform energy dispatch on the original energy allocation according to the actual energy consumption prediction value one or the actual energy consumption prediction value two, and output the actual energy demand allocation.

Images