CN-122022225-A - Industrial waste heat cascade utilization regulation and control method based on multi-source flow cooperation

CN122022225ACN 122022225 ACN122022225 ACN 122022225ACN-122022225-A

Abstract

The invention relates to the technical field of industrial waste heat utilization and intelligent regulation, in particular to an industrial waste heat cascade utilization regulation method based on multi-source flow cooperation, which comprises the steps of constructing a digital twin body to comprehensively deconstruct and simulate an industrial waste heat pipe network, accurately grasping the whole process of waste heat generation, transmission and consumption, prompting the system to distribute high-grade waste heat to high-demand users preferentially through a gradient stable rewarding function, reasonably utilizing medium-low temperature waste heat, avoiding waste of the waste heat, the cascade utilization of waste heat is realized, the overall utilization efficiency is remarkably improved, the punishment index of pipe network stability is fully considered in the decision-making process of the DRL algorithm, the excessive pressure fluctuation of the pipe network is avoided due to the regulation and control strategy, the DRL algorithm outputs the optimal regulation and control strategy covering multiple working conditions through trial and error learning in a digital twin environment, and meanwhile, the adaptability and stability of the algorithm under different scenes are ensured through covering the generalization capability of the applicable index evaluation strategy.

Inventors

LIU XI
YAN HU
SHEN GANHUA
CAI SHAN
CHEN HAO
WANG XINHUAN
HE WENLONG
CHEN QI

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: 20260512
Application Date: 20251205

Claims (9)

1. The industrial waste heat cascade utilization regulation and control method based on multi-source flow cooperation is characterized by comprising the following steps of: S1, deconstructing an industrial waste heat pipe network and constructing a digital twin body aiming at the industrial waste heat pipe network; Step S2, periodically collecting all the layered data of the industrial waste heat pipe network, importing the collected layered data into a digital twin body, further determining all possible working conditions, and extracting three-dimensional key characteristics of each possible working condition; Step S3, converting three-dimensional key features of each possible working condition into state vectors which can be identified by a DRL algorithm, defining an action space of the DRL algorithm, and defining gradient stability rewards; And S4, respectively importing the state vector of each possible working condition into a digital twin body, outputting an optimal regulation strategy of each possible working condition by the digital twin body based on a DRL algorithm, further acquiring a coverage applicability index of each optimal regulation strategy, marking the optimal regulation strategy with the maximum coverage applicability index as a waste heat cascade utilization final selection strategy, and applying the waste heat cascade utilization final selection strategy to carry out actual regulation.
2. The industrial waste heat cascade utilization regulation and control method based on multi-source flow coordination according to claim 1 is characterized by further determining all possible working conditions, specifically comprising the steps that after collected layered data are imported into a digital twin body, the digital twin body performs multiple virtual deductions based on the layered data, then a deduction proportion of each working condition is obtained, and when the deduction proportion of the working condition is higher than a proportion threshold value, the corresponding working condition is marked as the possible working condition.
3. The industrial waste heat cascade utilization regulation and control method based on multi-source flow coordination according to claim 1, wherein the three-dimensional key features comprise a source feature, a net feature and a load feature, the source feature comprises a waste heat grade classification feature, an output stability feature and a heat source equipment state feature, the net feature comprises a pipe network bearing capacity feature, a pressure stability feature, a temperature distribution and a heat loss feature, and the load feature comprises a demand grade matching feature, a demand gap and a stability feature.
4. The industrial waste heat cascade utilization regulation method based on multi-source flow coordination according to claim 1, wherein the gradient stabilization rewards are defined as follows: Wherein, the method comprises the steps of, Is an index of the gradient utilization of the waste heat, The penalty index for the stability of the pipe network, Is the gradient utilization coefficient of the waste heat, Punishment coefficients are used for pipe network stability.
5. The industrial waste heat cascade utilization regulation and control method based on multi-source flow coordination according to claim 4, wherein the waste heat gradient utilization index is as follows Wherein, the method comprises the steps of, On the basis of the recovery efficiency of the waste water, Is the grade matching coefficient.
6. The multi-source flow cooperation-based industrial waste heat cascade utilization regulation and control method according to claim 4, wherein the grade matching coefficient is as follows ; Is a high-temperature waste heat ratio, Is a medium temperature waste heat ratio, Is a low-temperature waste heat ratio, Is of high Wen Pipei coefficient, Is of a medium Wen Pipei coefficients, Is a low-Wen Pipei coefficient, and the coefficient is high, 。
7. The multi-source flow collaboration-based industrial waste heat cascade utilization regulation and control method as claimed in claim 4, wherein the pipe network stability penalty index is that of Wherein, the method comprises the steps of, Is the base pressure deviation; Is a critical node risk; Wherein, the method comprises the steps of, For the weight of the node, , Is the normalized absolute value of the pressure deviation of node i, ; For the real-time measured pressure of node i, Is the rated pressure of the node i.
8. The industrial waste heat cascade utilization regulation and control method based on multi-source flow coordination of claim 1, wherein the acquisition flow of coverage applicability index of the optimal regulation and control strategy is characterized by selecting an optimal regulation and control strategy, respectively introducing the optimal regulation and control strategy into a digital twin body with each possible working condition, and acquiring the optimal regulation and control strategy in the digital twin body with each possible working condition And further calculating and obtaining an average rewarding fluctuation index and an average gradient stabilizing rewarding index, calculating the difference value between the average gradient stabilizing rewarding index and the average rewarding fluctuation index, and calculating the coverage applicability index of the optimal regulation strategy.
9. The industrial waste heat cascade utilization regulation and control method based on multi-source flow coordination according to claim 8, wherein all gradient stable rewards are compared in pairs, absolute difference value calculation is carried out on the two compared gradient stable rewards, a reward fluctuation index is calculated, summation average value calculation is carried out on all the reward fluctuation indexes, average reward fluctuation index is calculated, summation average value calculation is carried out on all the gradient stable rewards, and average gradient stable reward index is calculated.

Description

Industrial waste heat cascade utilization regulation and control method based on multi-source flow cooperation Technical Field The invention relates to the technical field of industrial waste heat utilization and intelligent regulation, in particular to an industrial waste heat cascade utilization regulation method based on multi-source flow cooperation. Background In the industrial production process, a large amount of waste heat resources are not fully and reasonably utilized, and the problem that the waste heat grade is not matched with the terminal requirements exists. High-grade waste heat is often used in low-value scenes due to improper distribution, and the low-grade waste heat is easily discharged directly due to the limitation of the utilization technology, so that energy waste is caused. Meanwhile, the industrial waste heat pipe network is complex in structure and covers a plurality of links such as heat source, transmission, load and control, and the traditional regulation and control mode is difficult to accurately grasp dynamic changes of all links, so that waste heat utilization efficiency is low. The stable operation of the industrial waste heat pipe network faces a plurality of challenges, and the problems of pressure fluctuation, key node risks and the like of the pipe network are outstanding. The traditional regulation and control depend on manual experience or simple control logic, is difficult to cope with the dynamic change of working conditions, is easy to cause the conditions of unbalanced pipe network pressure, overlarge heat transmission loss and the like, not only affects the stability of waste heat supply, but also can cause equipment failure or potential safety hazard, and restricts the continuity and safety of industrial production. Along with the promotion of industrial intellectualization, the dual requirements of efficient utilization of waste heat and stable regulation and control of a pipe network cannot be met by simply relying on the traditional technology. The prior art has the defects in the aspects of working condition identification, intelligent decision and the like, lacks the capability of accurate simulation and dynamic optimization of the whole flow of a waste heat system, and is difficult to realize the cooperative regulation and control of waste heat cascade utilization and pipe network safe operation, and a technical scheme for fusing advanced modeling and intelligent algorithm is needed to break through the bottlenecks. Disclosure of Invention Aiming at the defects existing in the prior art, the invention aims to provide an industrial waste heat cascade utilization regulation and control method based on multi-source flow cooperation. In order to achieve the above purpose, the present invention provides the following technical solutions: industrial waste heat cascade utilization regulation and control method based on multi-source flow cooperation comprises the following steps: S1, deconstructing an industrial waste heat pipe network and constructing a digital twin body aiming at the industrial waste heat pipe network; Step S2, periodically collecting all the layered data of the industrial waste heat pipe network, importing the collected layered data into a digital twin body, further determining all possible working conditions, and extracting three-dimensional key characteristics of each possible working condition; Step S3, converting three-dimensional key features of each possible working condition into state vectors which can be identified by a DRL algorithm, defining an action space of the DRL algorithm, and defining gradient stability rewards; And S4, respectively importing the state vector of each possible working condition into a digital twin body, outputting an optimal regulation strategy of each possible working condition by the digital twin body based on a DRL algorithm, further acquiring a coverage applicability index of each optimal regulation strategy, marking the optimal regulation strategy with the maximum coverage applicability index as a waste heat cascade utilization final selection strategy, and applying the waste heat cascade utilization final selection strategy to carry out actual regulation. Further, all possible working conditions are further determined, specifically, after the collected layered data are imported into the digital twin body, the digital twin body performs multiple virtual deductions based on the layered data, then the deduction proportion of each working condition is obtained, and when the deduction proportion of the working condition is higher than a proportion threshold value, the corresponding working condition is marked as the possible working condition. Further, the three-dimensional key features comprise a source feature, a net feature and a load feature, wherein the source feature comprises a waste heat grade grading feature, an output stability feature and a heat source equipment state feature, the net feature compr