CN-121998213-A - Multi-energy-flow comprehensive energy management decision-making method for industrial park

CN121998213ACN 121998213 ACN121998213 ACN 121998213ACN-121998213-A

Abstract

The invention relates to the technical field of energy management, in particular to a multi-energy flow comprehensive energy management decision-making method for an industrial park. The method comprises the steps of collecting industrial waste heat data of a plurality of industrial waste heat generation sources in an industrial park, performing quality classification processing on the industrial waste heat data by using a K-means algorithm, determining the quality category of the industrial waste heat corresponding to each industrial waste heat generation source, constructing an energy transfer network model based on the industrial waste heat generation sources of the same quality category, selecting a path node to be selected for generating an energy transfer path from the energy transfer network model by using a random walk path generation algorithm, generating the energy transfer path of the industrial waste heat by adopting a rough combination mode for the path node to be selected, determining the utilization path of the industrial waste heat based on the energy transfer path, and forming a multi-energy collaborative heat and power combined system based on the utilization path, so that the utilization efficiency of the waste heat can be improved, the energy management efficiency can be improved, and the energy consumption can be reduced.

Inventors

YAN HU
SHEN GANHUA
LIU XI
LI JIANGQUAN
CAI SHAN
CHEN HAO
CHEN QI
WANG FAN

Assignees

新疆德润热力有限公司
新疆德润经济建设发展有限公司

Dates

Publication Date: 20260508
Application Date: 20251226

Claims (8)

1. The industrial park multi-energy flow comprehensive energy management decision-making method is characterized by comprising the following steps of: Collecting industrial waste heat data of a plurality of industrial waste heat generation sources in an industrial park, performing quality classification processing on the industrial waste heat data by using a K-means algorithm, and determining the quality category of the industrial waste heat corresponding to each industrial waste heat generation source; Constructing an energy transfer network model based on industrial waste heat generation sources of the same quality class, and selecting a path node to be selected for generating an energy transfer path from the energy transfer network model by utilizing a random walk path generation algorithm; generating an energy transfer path of industrial waste heat by adopting a rough combination mode for the path nodes to be selected; and determining an utilization path of the industrial waste heat based on the energy transmission path, and forming a multi-energy-source coordinated combined heat and power cooling system based on the utilization path.
2. The industrial park multi-energy flow integrated energy management decision method of claim 1, wherein the industrial waste heat data comprises waste heat temperature, waste heat medium pressure, waste heat medium flow and waste heat available heat.
3. The industrial park multi-energy flow comprehensive energy management decision-making method according to claim 1, wherein the quality classification processing is performed on the industrial waste heat data by using a K-means algorithm, and the determination of the industrial waste heat quality class corresponding to each industrial waste heat generation source specifically comprises the following processes: Step one, recording each industrial waste heat data as a sample , The value of (2) is N is the number of industrial waste heat generation sources, and k clustering centers are initialized randomly K is the number of preset clustering centers, the value of k is set according to the quality category of industrial waste heat, The value of (2) is ; Step two, calculating a sample The clustering center to be attributed is to calculate Euclidean distance between the clustering center and k clustering centers The calculation formula is as follows: wherein, the method comprises the steps of, Represents the first The value of the mth dimensional feature of the sample, Represent the first The M-th dimension feature of each cluster center is valued, M represents the total number of features, and after the distance between each sample and each cluster center is calculated, each sample is classified into a corresponding nearest cluster, wherein the cluster is the cluster center; Step three, calculating the deviation between the sample and the center of the cluster where the sample is located The calculation formula is as follows: ; updating a clustering center and repartitioning cluster attributions of all samples: wherein, the method comprises the steps of, Representing the updated first cluster center position, Representing belonging to a cluster Is used for the measurement of the sample of (a), Representing clusters of clusters The second and third steps are executed to calculate the new cluster division result and the new deviation ; Repeating the second to fourth steps, and iteratively updating the cluster division of the sample until the cluster deviation converges, namely the deviation difference value before and after the iterative updating Less than a given deviation threshold Determining a cluster corresponding to the industrial waste heat data of each industrial waste heat generation source, wherein the cluster is a quality class, and the iteration termination condition is as follows: wherein, the method comprises the steps of, Is the first A number of iterations of the process are performed, Is the deviation threshold for the iteration to terminate.
4. The industrial park multi-energy flow comprehensive energy management decision-making method according to claim 1, wherein the construction of the energy transfer network model based on the same quality class of industrial waste heat generation sources specifically comprises the following processes: Defining an energy transfer network model as , wherein, The method is a node set consisting of industrial waste heat generation sources with the same quality class, wherein the industrial waste heat generation sources are one node and the node set , In the form of a set of edges, one edge being represented in the form of a directed tuple, i.e 。
5. The industrial park multi-energy flow comprehensive energy management decision-making method according to claim 4, wherein selecting the candidate path node for generating the energy transfer path from the energy transfer network model by using the random walk path generation algorithm specifically comprises the following processes: Initializing random walk parameters, determining a starting node of random walk, and selecting the starting node from a node set V, wherein the random walk parameters comprise a walk number threshold value and a walk probability transition matrix updating period, and the starting node is determined from the node set V by random sampling or a mode of designating the node; using a random walk path generation algorithm to start from a starting node, carrying out random walk according to a directional connection relation of edges, and selecting a next node to be walked to arrive according to the edge condition of the current node when each step of walk is carried out and the preset probability; In the random walk process, a node sequence passing through the walk is recorded, the node sequence is determined to be a node sequence of a path to be selected of the energy transmission path, and the selection of the node of the path to be selected is completed.
6. The industrial park multi-energy flow comprehensive energy management decision-making method according to claim 5, wherein the calculation mode of the preset probability specifically comprises the following processes: For the current node Traversing all edges, counting and Total number of connected edges Calculating the slave Transfer to Probability of (2) , Wherein, the method comprises the steps of, Is a side Is used for the weight of the (c), Is a side The weights are pre-assigned based on historical flow and heat loss factors of the industry Yu Feire as it passes on the corresponding edge.
7. The industrial park multi-energy flow comprehensive energy management decision-making method according to claim 1, wherein the energy transfer path for generating industrial waste heat by adopting a rough combination mode for the path nodes to be selected specifically comprises the following processes: Step one, setting For coarse combination parameters, for each Coarse combination is carried out on the path nodes to be selected according to the sampling sequence of the sampling interval values, so that a plurality of coarse combination groups are obtained; Step two, the first node and the second node of each coarse combination group form a reference vector Will leave the first The individual node Individual node composition vectors , wherein, Respectively calculate And Angle therebetween Wherein the angle is The calculation formula of (2) is as follows: ; Step three, will be Angle to threshold value Comparing, if Of individual points Greater than a threshold angle Will be arranged at the first The nodes in front of the individual nodes are subdivided into groups to be arranged in the first Repeating the second and third steps by the node in front of each node to finish the subdivision of the coarse combination, and marking each subdivided group as a segmented path; and step four, combining the segmented paths into an energy transmission path of industrial waste heat according to the sampling sequence.
8. The industrial park multi-energy flow comprehensive energy management decision-making method according to claim 1, wherein the method is characterized in that the utilization path of the industrial waste heat is determined based on the energy transmission path, and a multi-energy cooperative combined heat and power system is formed based on the utilization path, and specifically comprises the following processes: The utility model provides an integration multiple energy input, including electric wire netting electric power, gas, wind power generation, photovoltaic power generation and fused salt heat accumulation renewable energy and energy storage mode, forms the thermoelectric cold allies oneself with the confession system of the collaborative input of multipotency source, and the thermoelectric cold allies oneself with the confession system of multipotency source cooperation includes: The heat supply part is used for conveying the converted heat energy to the client through the thermodynamic ring network so as to meet the heat requirement of a user; The cold supply part is used for converting heat energy into cold energy by utilizing the absorption refrigerating unit and providing the cold energy for the client to meet the cold requirement; the electricity supply part converts heat energy into electric energy through a cogeneration unit, and ensures stable supply of electric power by combining electric power input of a power grid, wind power generation and photovoltaic power generation; and the steam supply part is used for generating steam through the evaporator equipment and meeting the requirements of clients on the steam.

Description

Multi-energy-flow comprehensive energy management decision-making method for industrial park Technical Field The invention relates to the technical field of energy management, in particular to a multi-energy flow comprehensive energy management decision-making method for an industrial park. Background In the industrial production process, a large amount of waste heat, such as waste heat of a sewage source heat pump, waste heat of steam/exhaust gas, waste heat of high-temperature flue gas and the like, is generated. The waste heat contains huge energy, if the waste heat is effectively utilized, the energy consumption of a park can be reduced, the dependence on the traditional energy is reduced, the emission of greenhouse gases can be obviously reduced, and the aims of energy conservation and emission reduction are fulfilled. In the prior art, industrial parks have a plurality of problems in terms of waste heat utilization. On the one hand, due to uneven quality of the waste heat, the waste heat from different sources has larger difference in parameters such as temperature, flow and the like, so that the traditional waste heat utilization mode is single, the waste heat quality is not accurately classified and utilized in a targeted manner, partial waste heat cannot be fully utilized, even the waste heat is directly discharged, and energy waste is caused. On the other hand, the energy transfer path of the waste heat is not reasonably planned, and the waste heat utilization rate is further low. Therefore, there is a need for a multi-energy comprehensive energy management decision method for industrial parks to solve the above problems. Disclosure of Invention The invention aims to solve the technical problem that the industrial park has low utilization efficiency of waste heat of the industrial park. The method for making the decision on the multi-energy-flow comprehensive energy management of the industrial park comprises the following steps: Collecting industrial waste heat data of a plurality of industrial waste heat generation sources in an industrial park, performing quality classification processing on the industrial waste heat data by using a K-means algorithm, and determining the quality category of the industrial waste heat corresponding to each industrial waste heat generation source; Constructing an energy transfer network model based on industrial waste heat generation sources of the same quality class, and selecting a path node to be selected for generating an energy transfer path from the energy transfer network model by utilizing a random walk path generation algorithm; generating an energy transfer path of industrial waste heat by adopting a rough combination mode for the path nodes to be selected; and determining an utilization path of the industrial waste heat based on the energy transmission path, and forming a multi-energy-source coordinated combined heat and power cooling system based on the utilization path. Further, the industrial waste heat data includes industrial waste heat data including waste heat temperature, waste heat medium pressure, waste heat medium flow, and waste heat available heat. Further, the quality classification processing is carried out on the industrial waste heat data by using a K-means algorithm, and the determination of the industrial waste heat quality class corresponding to each industrial waste heat generation source specifically comprises the following steps: Step one, recording each industrial waste heat data as a sample ,The value of (2) isN is the number of industrial waste heat generation sources, and k clustering centers are initialized randomlyK is the number of preset clustering centers, the value of k is set according to the quality category of industrial waste heat,The value of (2) is; Step two, calculating a sampleThe clustering center to be attributed is to calculate Euclidean distance between the clustering center and k clustering centersThe calculation formula is as follows: wherein, the method comprises the steps of, Represents the firstThe value of the mth dimensional feature of the sample,Represent the firstThe M-th dimension feature of each cluster center is valued, M represents the total number of features, and after the distance between each sample and each cluster center is calculated, each sample is classified into a corresponding nearest cluster, wherein the cluster is the cluster center; Step three, calculating the deviation between the sample and the center of the cluster where the sample is located The calculation formula is as follows: ; updating a clustering center and repartitioning cluster attributions of all samples: wherein, the method comprises the steps of, Representing the updated first cluster center position,Representing belonging to a clusterIs used for the measurement of the sample of (a),Representing clusters of clustersThe second and third steps are executed to calculate the new cluster division result and the new deviat