Heavy industry facility load optimization scheduling method and system based on partition and layering

Abstract

The invention relates to the technical field of power load management, in particular to a heavy industrial facility load optimization scheduling method and system based on partition and layering. Comprising the following steps: dividing load equipment into a plurality of functional areas according to a production process flow and an equipment physical position of the industrial facility; performing priority layering on the equipment in each functional area based on the adjustable potential of the equipment; according to the partitioning and layering results, constructing and solving a multi-target optimization scheduling model taking embedded layering constraint as a core, and generating a load scheduling instruction; and executing the scheduling instruction, and collecting a system operation state in real time to perform dynamic feedback correction on the instruction. According to the method provided by the invention, the refined and differentiated sorting of the load resources is realized, so that the subsequent optimal scheduling has a scientific and reliable basis, and the management mode is closer to the actual industrial production.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power load management, and particularly relates to a heavy industrial facility load optimization scheduling method and system based on zoning and layering. BACKGROUND

[0002] Building a new power system with new energy as the main body is the core measure to achieve the national "carbon peak and carbon neutral" strategic goal. With the continuous rise of the proportion of wind power, photovoltaic and other volatile renewable energy installations, the power system is facing huge balancing pressure and flexible regulation demand. The traditional "source follows load" balancing mode is difficult to continue, and it is urgent to tap the regulation potential from the demand side to ensure the safe, stable and economic operation of the power grid. The industrial sector, as the absolute main body of electricity consumption, especially heavy industrial facilities such as steel, chemical industry, cement, etc., has a huge amount of electricity-consuming equipment inside, which constitutes a large-scale and highly valuable adjustable load resource pool.

[0003] Therefore, the national policy level continues to promote the improvement of the demand side response mechanism, aiming to guide industrial users to actively participate in grid regulation through price signals and market incentives. However, in the current technical practice level, there are still significant bottlenecks and challenges in the scheduling and management of industrial loads. The existing mainstream load management and scheduling technology generally adopts a "black box" and "one-size-fits-all" extensive mode. Specifically, the grid dispatching agency or load aggregator usually regards the entire complex industrial plant as a single, homogeneous load node, only focusing on the total power change at the grid access point, and cannot deeply perceive and coordinate the highly heterogeneous equipment group and its complex production logic inside.

[0004] This extensive management mode is due to the neglect of the inherent differences in industrial production systems. In a typical energy-intensive factory (such as a steel joint venture), different production units and equipment differ greatly in functional criticality, operating constraints and adjustable potential. For example, the core smelting equipment that guarantees the continuous operation of blast furnaces and converters has very high rigidity of load, and any fluctuation may cause great production safety risks and economic losses; while the auxiliary or non-continuous production equipment such as water pumps, air compressors and some rolling mills in the plant have considerable load adjustment flexibility.

[0005] The equivalence of the two devices with different characteristics leads to two irreconcilable contradictions: one is that the response potential is seriously wasted. In order to absolutely guarantee the safety of the core production, the enterprise often adopts an extremely conservative strategy when responding to the grid command, and only a small number of auxiliary devices known to be safe are mobilized, so that a large amount of adjustable potential hidden in the secondary core and auxiliary production links cannot be used to provide effective flexibility support for the system. The second is that the production safety is threatened. If the load control command of the power grid is too severe, and the enterprise lacks fine internal load allocation and decision-making tools, it may be forced to disturb the operation of the core production equipment, and the chain reaction and economic loss caused by it far exceeds the benefits obtained by participating in demand response.

[0006] Therefore, there is a huge technical gap between the macro grid dispatching demand and the micro industrial production constraint in the prior art system. The traditional load aggregation and dispatching method cannot effectively balance the two core goals of "guaranteeing the absolute safety of production" and "mining the economic value of response", which has become a key obstacle to the large-scale, high-quality access of industrial load resources to new power systems. An innovative technical solution is needed that can penetrate the "black box", deeply identify, structure, and differentiate the heterogeneous load resources within the industrial facility. SUMMARY

[0007] The present application provides a partitioned and layered heavy industrial facility load optimization scheduling method and system to solve the technical problems in the background art.

[0008] A partitioned and layered heavy industrial facility load optimization scheduling method, comprising the following steps: step S1: dividing the load devices into multiple functional areas according to the production process and physical location of the industrial facility; step S2: prioritizing and layering the devices in each functional area based on their adjustable potential; step S3: constructing and solving a multi-objective optimization scheduling model with embedded hierarchical constraints based on the partitioning and layering results, and generating load scheduling instructions; step S4: executing the scheduling instructions and collecting system operation status in real time to dynamically feedback and correct the instructions.

[0009] Preferably, the step S1 specifically comprises: dividing the entire set of devices into , each subset corresponding to a functional area , satisfying: .

[0010] ​​Preferably, the step S2 is implemented by using the TOPSIS method, comprising: constructing a decision matrix containing response capacity, response time, response duration, response cost and reliability index of the equipment; performing standardization processing on the decision matrix; calculating positive ideal solution and negative ideal solution of each index; calculating Euclidean distance of each equipment from the positive ideal solution and the negative ideal solution; calculating relative closeness of each equipment according to the distance . The priority of all equipment is sorted according to the relative closeness.

[0011] Preferably, the objective function of the multi-objective optimization scheduling model constructed in the step S3 is:

[0012]

[0013]

[0014] In the formula, C is total power consumption cost, is scheduling pressure cost, is compensation income, is actual operation load of the equipment at time t, is electricity price at time t, is scheduling pressure cost coefficient of the equipment at time t, is regulation capacity of the equipment at time t, is compensation income coefficient of the equipment at time t. Preferably, the constraint condition of the multi-objective optimization scheduling model comprises: equipment maximum response constraint: ; regional capacity constraint: ; equipment minimum reference load constraint: ; hierarchical scheduling rule constraint: . Cross-zone load transfer constraint: . Preferably, the dynamic feedback mechanism in the step S4 corrects the scheduling instruction by the following formula:

[0015] In the formula, is the regulation instruction of the next time t+1, and is the actual operation load of the equipment at time t.

[0016] Preferably, the dynamic feedback mechanism in the step S4 corrects the scheduling instruction by the following formula: In the formula, is the regulation instruction of the next time t+1, and is the actual operation load of the equipment at time t.​​​​​​​​ a regulation instruction for a current moment, a system state deviation for the current moment, a feedback gain coefficient.

[0017] The application further discloses a heavy industrial facility load optimization scheduling system based on zoning and layering, which is used for realizing the method and comprises a zoning management module, a layering management module, an optimization scheduling module and a dynamic feedback module.

[0018] Preferably, the zoning management module is specifically used for dividing the device set into a plurality of mutually disjoint device subsets. Preferably, the layering management module specifically comprises a decision matrix construction unit, a standardization processing unit, an ideal solution calculation unit, a distance calculation unit and a priority sorting unit. .

[0019] Preferably, the layering management module specifically comprises a decision matrix construction unit, a standardization processing unit, an ideal solution calculation unit, a distance calculation unit and a priority sorting unit. Preferably, the optimization scheduling module comprises a target function construction unit and a constraint processing unit.

[0020] Preferably, the optimization scheduling module comprises a target function construction unit and a constraint processing unit. Preferably, the constraint processing unit is used for processing device maximum response constraints, region capacity constraints, device minimum reference load constraints and hierarchical scheduling rule constraints.

[0021] The technical scheme provided by the application has the beneficial effects that: The application combines the zoning based on production processes and the layering based on adjustable potential for the first time, and constructs a two-dimensional management framework. The framework changes the originally heterogeneous, complex and disordered industrial device group into a structured resource pool with clear levels, distinct characteristics and accurate scheduling. This enables the scheduling system to deeply perceive the internal logic of industrial production, realize the differentiated and refined management and scheduling of different importance and different characteristics of loads, and completely changes the extensive management mode of traditional methods regarding the entire factory as a black box or a single node.

[0022] ​The application embeds the device priority as a rigid constraint in the multi-objective optimization scheduling model, which ensures the stable operation of high-priority core devices at the algorithm level. When the model globally optimizes for economy, it automatically prioritizes the use of low-priority, high-flexibility auxiliary devices to respond to grid demand, thereby eliminating the risk of affecting core production safety due to scheduling instructions at the root. The application successfully balances the two traditionally difficult-to-reconcile goals of "ensuring production safety" and "pursuing response benefits."

[0023] By scientifically evaluating and ranking the adjustable potential of devices through multi-attribute decision-making methods such as TOPSIS, the application can comprehensively and deeply tap the flexible load resources hidden in various production links, avoiding the waste of response potential. This enables enterprises to release much larger adjustable capacity than traditional methods under the premise of ensuring safety, thereby more efficiently and reliably participating in grid demand response projects and obtaining more substantial economic compensation benefits.

[0024] The solution provided by the application is mainly based on the innovation of software algorithms and system architecture, without the need for large-scale hardware modification or interruption of production of existing industrial production lines. The system can be integrated as an advanced application module into the existing energy management system (EMS) or production execution system (MES) of the enterprise, obtaining information and issuing instructions through data interfaces, greatly reducing the implementation threshold and cost. This non-intrusive feature makes it highly valuable and promising for application in various energy-intensive enterprises such as steel, chemical, and cement.

[0025] By providing large-scale, fine-grained, and highly reliable industrial adjustable load resources for the grid, the application effectively improves the overall flexibility and regulation capacity of the power system. This helps to smooth the random volatility of renewable energy generation, alleviates the system peak-valley difference pressure, and has important positive significance for ensuring the safe and stable operation of new power systems, accelerating energy structure transformation, and serving the national "double carbon" strategic goal. BRIEF DESCRIPTION OF DRAWINGS

[0026] Figure 1 Schematic diagram of production process and main equipment partitioning for steel enterprises; Figure 2 Schematic diagram of dynamic process of load response; Figure 3 Schematic diagram of hierarchical classification and sorting; Figure 4 Schematic diagram of partitioning and layering matching method framework. DETAILED DESCRIPTION

[0027] In order to make the purpose, technical scheme and advantages of the present application more clear, the present application is further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application.

[0028] As shown in Figure 1 , a heavy industrial facility load optimization scheduling method based on zoning and layering includes the following steps: Step A. Zoning and layering modeling of industrial load resources First, define the full set of all schedulable load devices in the industrial facility as , where N is the total number of devices. According to the production process, define a set of functional areas . Divide the full set to get the device subset corresponding to each functional area , where . At the same time, it is necessary to ensure that the union of all device subsets Figure 1 is equal to the full set E, and the device subsets corresponding to different functional areas are mutually exclusive. That is, formula (1) is satisfied. The mapping result is used as the input of the subsequent layering modeling to ensure that the scheduling model can be based on the actual production logic.

[0029] (1) Step B. Adjustable potential evaluation and layering model based on TOPSIS method On the basis of traditional linear weighting model, introduce a more mature multi-attribute decision-making theory-TOPSIS method to scientifically, systematically and quantitatively evaluate and sort the adjustable potential of each device. Through this method, the priority of the response capacity of multi-load resources can be divided, and the decision basis for subsequent zoning and layering management can be provided.

[0030] When applying the TOPSIS method, a decision matrix is first constructed, which includes multiple device-related evaluation indexes. Suppose there are N devices and evaluation indexes. Among them, the evaluation indexes include two aspects: one is the controllability index, which specifically includes response capacity, response time and response duration (see formulas (2) to (4)); the other is the economic index, which covers response cost and reliability (see formulas (5) to (6)).

[0031] Controllability index: (2) (3) (4) Economic index: (5) (6) where, denotes the initial load level before the response of the industrial adjustable equipment starts; denotes the minimum operating power of the equipment; denotes the time of issuing the control command; denotes the full response time; denotes the end time of the response; denotes the difference in power caused by the load adjustment; denotes the compensation price paid by the power grid to the user; denotes other additional costs such as maintenance and equipment wear and tear; is an uncertainty parameter. is a reliability factor. Figure 2 is a dynamic process diagram of load response.

[0032] According to the numerical values of each equipment in each evaluation index, the following decision matrix X can be constructed: (7) where, and respectively represent the evaluation values of the i-th equipment in the adjustment potential and the economy.

[0033] In order to eliminate the influence of dimension on the calculation results, the decision matrix needs to be standardized. The standardized matrix is calculated by the following formula: (8) where, is the element in the i-th row and j-th column of the original decision matrix, is the element of the standardized matrix.

[0034] In the TOPSIS method, the positive ideal solution (PIS) and the negative ideal solution (NIS) represent the optimal and worst values of each index, respectively. Specifically: the positive ideal solution : the maximum value (for positive indexes such as adjustment potential) and the minimum value (for negative indexes such as economic index) of each index; the negative ideal solution : the minimum value (for positive indexes) and the maximum value (for negative indexes) of each index. Therefore, the calculation formula of the positive ideal solution and the negative ideal solution is: (9) (10) In the calculation process, the Euclidean distance of each device to the positive ideal solution and the negative ideal solution needs to be calculated. The specific calculation formula is as follows: Distance to the positive ideal solution: (11) Distance to the negative ideal solution: (12) In the formula, and are the values of the jth evaluation index in the positive ideal solution and the negative ideal solution, respectively.

[0035] The relative closeness is calculated as the basis for the comprehensive advantages and disadvantages of the device. The relative closeness is calculated by the distance to the positive ideal solution and the negative ideal solution, and the formula is as follows: (13) According to the calculated relative closeness , all devices are sorted. In the sorting result, the device with a higher value indicates that its comprehensive performance in adjustment potential and economy is better, and thus it is considered as the preferred device.

[0036] Through the above steps, the complex group of devices in the region is further refined into a clear priority structure. The hierarchical modeling result provides a reliable basis for subsequent scheduling tasks. As shown in FIG. 3, the TOPSIS method can comprehensively consider the controllability and economy, realize the hierarchical sorting of devices, and form a clear hierarchical load management structure.

[0037] Step C. Scheduling tasks and load allocation On the basis of the aforementioned load priority sorting, the main task of this step is to ensure the maximization of the economy and stability of the power system under the premise of meeting the production demand through regional allocation strategy and device scheduling. A closed-loop control system of “response task → zoning and layering → dynamic feedback” is constructed. The core of the strategy is the scheduling logic of “main first and auxiliary second, layering and gradual progress”; and the dynamic matching mechanism of “state perception + strategy reflection” (see Figure 4 ).

[0038] The closed-loop control system of “response task → zoning and layering → dynamic feedback” corresponds to the following optimization framework at the modeling level: wherein the objective function is used to reflect the comprehensive requirements of the response task on the economy and stability; the zoning and layering results are converted into capacity and priority constraints; and the dynamic feedback mechanism is realized through the real-time deviation correction formula.

[0039] The target function is set as follows: (14) (15) (16) (17) wherein, is the total power consumption cost, is the dispatch pressure cost, is the compensation benefit, is the equipment is the actual operation load at time , is the electricity price at time , is the dispatch pressure cost coefficient of equipment at time , is the adjustment capacity of equipment at time , and is the compensation benefit coefficient of equipment at time .

[0040] Equipment operation constraints (1) Equipment maximum response constraint The adjustment range of a single equipment cannot exceed its own maximum response capability, i.e., formula (18) is satisfied.

[0041] (18) (2) Regional capacity constraint The total load of equipment in a functional area cannot exceed the upper limit of the capacity of the area, i.e., formula (19) is satisfied.

[0042] (19) (3) Priority rigid constraint Each equipment needs to maintain a minimum baseline load to ensure the stable operation of the key production link, i.e., formula (20) is satisfied.

[0043] (20) Dispatching strategy constraints (1) Hierarchical dispatching rule The adjustment range of equipment at different levels needs to follow the hierarchical logic of "minimum in the core layer and maximum in the auxiliary layer", i.e., formula (21) is satisfied.

[0044] (21)​​​​​ (2) Cross-zone load transfer constraint When a certain area is short of load, it can be balanced through cross-zone allocation, but the transfer amount is limited by the supply margin and demand gap . That is, formula (22) is satisfied. (22) Dynamic feedback mechanism This mechanism dynamically corrects the dispatching results by collecting the system running state in real time, to ensure that the dispatching scheme can adapt to the fluctuations of the power grid and production conditions. Let the system state deviation be , then the dispatching instruction update formula is: (23) In the formula, is the feedback gain coefficient.

[0045] Through the above dispatching strategy, combined with the objective function and constraint conditions, the stability of the production process and the economy of power dispatching are ensured. All device load dispatching is carried out within the limit of response capacity, while considering multiple factors such as power cost, dispatching pressure, compensation benefit, etc., to maximize the overall benefit of the system.

[0046] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for optimizing load scheduling of heavy industrial facilities based on partitioning and hierarchical division, characterized in that, The steps include: Step S1: Divide the load equipment into multiple functional areas according to the production process flow and physical location of the equipment in the industrial facility; Step S2: Prioritize the equipment in each functional area based on the adjustable potential of the equipment. Step S3: Based on the partitioning and layering results, construct and solve a multi-objective optimization scheduling model with embedded layering constraints as the core, and generate load scheduling instructions; Step S4: Execute the scheduling instructions and collect the system operating status in real time to dynamically correct the instructions.

2. The method according to claim 1, characterized in that, Step S1 specifically includes: assembling the entire device set Divided into A set of non-overlapping devices Each subset corresponds to a functional area. ,satisfy: .

3. The method according to claim 1, characterized in that, Step S2 is specifically implemented using the TOPSIS method, including: constructing a decision matrix containing indicators such as device response capacity, response time, response duration, response cost, and reliability; standardizing the decision matrix; calculating the positive and negative ideal solutions for each indicator; calculating the Euclidean distance between each device and the positive and negative ideal solutions; and calculating the relative proximity of each device based on the distances. Based on the aforementioned relative proximity Prioritize all devices.

4. The method according to claim 1, characterized in that, The objective function of the multi-objective optimization scheduling model constructed in step S3 is: in, ; ； ； In the formula, It is the total cost of electricity consumption. It is the cost of scheduling pressure. It is compensation income. For equipment At any moment Actual operating load For a moment Electricity price, It is equipment At any moment The scheduling pressure cost coefficient, equipment At any moment Adjustable capacity, equipment At any moment The compensation benefit coefficient.

5. The method according to claim 4, characterized in that, The constraints of the multi-objective optimization scheduling model include: maximum device response constraint: Regional capacity constraints: Minimum base load constraint for equipment: Hierarchical scheduling rule constraints: ; Inter-regional load transfer constraints: .

6. The method according to claim 1, characterized in that, The dynamic feedback mechanism in step S4 modifies the scheduling instructions using the following formula: in, For the adjustment instructions in the next moment, The adjustment command for the current moment. The system state deviation at the current moment. This is the feedback gain coefficient.

7. A partitioned and hierarchical load optimization scheduling system for heavy industrial facilities, used to implement the method according to any one of claims 1 to 6, characterized in that, include: The partition management module is used to execute step S1, dividing the load equipment into multiple functional areas; The hierarchical management module is used to execute step S2, which prioritizes the devices. The optimization scheduling module is used to execute step S3, construct and solve the optimization model to generate scheduling instructions; the dynamic feedback module is used to execute step S4, dynamically correct the scheduling instructions.

8. The system according to claim 7, characterized in that, The partition management module is specifically used for: managing the entire set of devices Divided into A set of non-overlapping devices .

9. The system according to claim 7, characterized in that, The hierarchical management module specifically includes: a decision matrix construction unit for constructing an evaluation index decision matrix; a standardization processing unit for standardizing the decision matrix; an ideal solution calculation unit for calculating positive and negative ideal solutions; a distance calculation unit for calculating Euclidean distance; and a priority ranking unit for calculating relative proximity. And sort them.

10. The system according to claim 6, characterized in that, The optimization scheduling module includes: an objective function construction unit, used to construct the objective function. The constraint processing unit is used to process equipment maximum response constraints, regional capacity constraints, equipment minimum baseline load constraints, and hierarchical scheduling rule constraints.

Images