Centralized and distributed comprehensive demand response method and system of comprehensive energy system

Abstract

The invention discloses a centralized distributed comprehensive demand response method and system for a comprehensive energy system, and the method comprises the steps: obtaining the planned power consumption of each park, generating noise disturbance based on a differential privacy mechanism, and determining the response amount of each park through the peak regulation demand of a power grid and the planned power consumption after the noise disturbance is increased; taking a difference value between the planned power consumption of each park and the response quantity as a power consumption reference value of each park participating in response; according to the difference value between the electricity utilization reference value of each park participating in response and the actual electricity utilization amount of each park, the reward and punishment cost of each park obtained from the integrated energy park operator is determined; taking the response quantity corresponding to the minimum value of the difference value between the operation cost and the reward and punishment cost of each park as the optimal response quantity of each park; the minimum error between the optimal response amount and the response amount of each park and the maximum income of the integrated energy park operator are taken as optimization objectives, and the demand response amount instruction of each park is determined, so that the problems of privacy disclosure, low participation intention and heavy calculation burden of the industrial park in the existing demand response mechanism are solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of energy demand response of smart grids. Specifically, it relates to a centralized and distributed integrated demand response method and system for an integrated energy system. Background Art

[0002] At present, the application of the integrated demand response mechanism is still in its initial stage, lacking a set of practical and effective integrated demand response mechanisms. In practical applications, although various methods have been proposed, due to the lack of a unified standard and coordination mechanism, the existing demand response schemes are difficult to implement comprehensively and effectively, especially in the case of involving residential and commercial users. The load distribution of residential and commercial users is relatively dispersed, the load of a single user is small, and the number of users is large. Therefore, when conducting demand response, it is difficult to coordinate the interests among individuals, the difficulty of aggregating overall flexible resources is large, and the scheduling efficiency is low. In addition, the demand response preferences of different types of users vary greatly, increasing the complexity of flexibility management and control. Therefore, the popularization and application of integrated demand response in these user groups face great challenges at the present stage. In contrast, industrial parks have relatively large and concentrated loads, and load regulation is relatively easy. Therefore, industrial parks, as the implementation objects of demand response, have the potential to implement integrated demand response. Industrial parks not only have strong demand response capabilities, but also usually involve the use of multiple energy sources such as electricity, heat, cold, and gas, forming a typical integrated energy system. In this multi-energy collaborative environment, the energy utilization in the park can be scheduled and respond to demand fluctuations more efficiently, thus better achieving the demand response goal. Based on this characteristic, the application of industrial parks in demand response has great advantages. However, although industrial parks have advantages in implementing integrated demand response, there are still some deficiencies in the existing technologies. The existing demand response mechanisms often operate in a centralized or distributed manner, lacking effective integration and optimization of the two, resulting in the potential in response efficiency, scheduling flexibility, and resource utilization not being fully explored. In addition, the energy system in the park involves multiple energy forms, and the existing scheduling methods do not fully consider the synergistic effects among various energy sources, thus affecting the overall energy management efficiency of the park. Summary of the Invention

[0003] To solve the deficiencies in the existing technology, the present invention provides a centralized and distributed integrated demand response method and system for an integrated energy system, which does not require a large amount of information provided by each industrial park, protects the privacy of each industrial park, thereby improving the enthusiasm of industrial parks to participate in demand response, and decentralizes more calculations to be completed by each park itself, reducing the requirements for the performance of the integrated energy park operator, reducing the solution time, and solving problems such as privacy leakage, low participation willingness, and heavy computational burden in the demand response mechanism of the existing technology.

[0004] The present invention adopts the following technical solution.

[0005] The present invention proposes a centralized and distributed integrated demand response method for a comprehensive energy system, including: Obtaining the planned electricity consumption of each park, generating noise perturbation based on the differential privacy mechanism, and determining the response amount of each park by using the peak shaving demand of the power grid and the planned electricity consumption after adding noise perturbation; taking the difference between the planned electricity consumption and the response amount of each park as the electricity consumption benchmark value for each park to participate in the response; Determining the reward and punishment costs obtained by each park from the comprehensive energy park operator according to the difference between the electricity consumption benchmark value for each park to participate in the response and the actual electricity consumption of each park; taking the response amount corresponding to the minimum value of the difference between the operating cost and the reward and punishment cost of each park as the optimal response amount of each park; Taking the minimum error between the optimal response amount and the response amount of each park and the maximum income of the comprehensive energy park operator as the optimization objective, and determining the demand response amount instruction of each park.

[0006] Preferably, obtaining the planned electricity consumption of each park, generating noise perturbation based on the differential privacy mechanism, and the park reports the planned electricity consumption after adding noise perturbation to the operator; Park i During the peak shaving period t The planned electricity consumption is , and the park locally Adds noise perturbation , to obtain the planned electricity consumption after adding noise perturbation = + ; Among them, generating noise perturbation based on the differential privacy mechanism, the probability distribution function of the noise perturbation Is shown by the following relational expression:

[0007] In the formula, Is the probability distribution function of the noise perturbation , Is the noise control parameter, .

[0008] Preferably, based on the planned electricity consumption After adding noise perturbation to each park and the peak shaving demand of the superior power grid, the peak shaving demand of the power grid is allocated to each park as the response amount of each park, satisfying the following relational expression:

[0009] In the formula, Is the park i During the peak shaving periodt The response volume, is the peak shaving demand of the power grid during the peak shaving period t and is the planned power consumption of the park i during the peak shaving period t after adding noise disturbance.

[0010] Preferably, the difference between the planned power consumption and the response volume of each park is used as the power consumption benchmark value for each park to participate in the response, and the following relationship is satisfied:

[0011] In the formula, is the power consumption benchmark value for the park i to participate in the response during the peak shaving period t and is the planned power consumption of the park i during the peak shaving period t .

[0012] Preferably, the park obtains corresponding penalties or rewards from the integrated energy park operator, and the following relationship is satisfied:

[0013] In the formula, when , is the reward given by the integrated energy park operator to the park i during the peak shaving period t , and when , is the penalty given by the integrated energy park operator to the park i during the peak shaving period t , is the power consumption of the park i during the peak shaving period t , is the unit price of the reward or penalty during the peak shaving period t , is the disturbance compensation factor of the park i during the peak shaving period t .

[0014] Preferably, the disturbance compensation factor satisfies the following relationship: .

[0015] Preferably, the optimization objective satisfies the following relationship:

[0016] In the formula, is the optimization objective, is the park i during the peak shaving periodt Optimal response volume For the park i During the peak shaving period t Response volume Is the income of the integrated energy park operator

[0017]

[0018]

[0019] Wherein Is the electricity sales income of the integrated energy park operator For the peak shaving period t Electricity sales unit price For the gas turbine during the peak shaving period t Output electric power

[0020] Wherein Is the heat sales income of the integrated energy park operator Is the heat sales unit price

[0021] Wherein Is the income of the integrated energy park operator from selling peak shaving volume For the peak shaving period t Unit price of selling peak shaving volume For the power grid during the peak shaving period t Peak shaving demand

[0022] Wherein For the gas turbine during the peak shaving period t Fuel cost For the gas turbine during the peak shaving period Output electric power 、 、 Are the cost coefficients corresponding to the output of unit electric power respectively For the gas turbine during the peak shaving period Output heat power 、 、 Are the cost coefficients corresponding to the output of unit heat power respectively

[0023] Wherein Is the compensation expenditure of the integrated energy park operator For the integrated energy park operator to the parki During the peak shaving period t Rewards or penalties

[0024] Preferably, when the error between the optimal response quantity and the response quantity of each park is minimized and the revenue of the integrated energy park operator is maximized, the demand response quantity command of each park satisfies the following relational expression:

[0025] In the formula is the park i During the peak shaving period t Demand response quantity command is the park i During the peak shaving period t Planned electricity consumption is the peak shaving demand of the power grid during the peak shaving period t t is the electric power output by the gas turbine during the peak shaving period

[0026] The present invention also proposes a centralized and distributed integrated demand response system for an integrated energy system, including: A park execution module, configured to obtain the planned electricity consumption of each park, generate noise perturbations based on the differential privacy mechanism, and determine the response quantity of each park by using the peak shaving demand of the power grid and the planned electricity consumption after adding noise perturbations; use the difference between the planned electricity consumption and the response quantity of each park as the electricity consumption benchmark value for each park to participate in the response; An operator execution module, configured to determine the reward and punishment costs obtained by each park from the integrated energy park operator according to the difference between the electricity consumption benchmark value for each park to participate in the response and the actual electricity consumption of each park; use the response quantity corresponding to the minimum value of the difference between the operating cost and the reward and punishment cost of each park as the optimal response quantity of each park;

[0027] A comprehensive execution module, configured to determine the demand response quantity command of each park with the optimization goal of minimizing the error between the optimal response quantity and the response quantity of each park and maximizing the revenue of the integrated energy park operator.

[0028] The beneficial effects of the present invention are at least as follows compared with the prior art. The method proposed by the present invention optimizes the response quantity and revenue of industrial parks and integrated energy park operators, ensures that industrial parks can actively participate in demand response while protecting privacy, and at the same time improves the economic benefits of integrated energy park operators. Significantly reduces the amount of external information that each industrial park needs to provide, and effectively protects the operation data and privacy of the park. Through a centralized-distributed combination method, the park completes data processing and demand response calculation internally, avoiding sharing sensitive data externally, thereby enhancing the security and willingness of the park to participate.

[0028] By decentralizing computing tasks to each park, the present invention reduces the computing requirements for the central operator and the load on the central system. This distributed computing method not only reduces the consumption of computing resources, but also shortens the system solving time, improves the response speed, and thus greatly enhances the overall operation efficiency. With the optimized centralized-distributed scheduling strategy, more flexible energy management can be achieved on the premise of ensuring the demand response effect, reducing the complexity of energy management within the park, while enhancing the park's ability and flexibility to participate in demand response. The park does not need to expose too many operation details and can independently control the computing and decision-making in the participation process, increasing the park's enthusiasm for participating in the demand response mechanism. Compared with the traditional centralized demand response mode, the present invention significantly improves the acceptance and coordination degree of the park to the system. Brief Description of the Drawings

[0029] Figure 1 It is a flowchart of the centralized and distributed integrated demand response method for the integrated energy system proposed by the present invention; Figure 2 It is the total electricity purchase volume of the industrial park aggregate after response in the embodiment of the present invention. Detailed Embodiment

[0030] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. The embodiments described in this application are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the spirit of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the protection scope of the present invention.

[0031] Based on the existing centralized-distributed architecture, the present invention proposes an improved demand response coordination method, which not only conducts collaborative optimization of upper-layer profit and lower-layer cost through a two-layer optimization model, but also introduces a privacy-protected data interaction mechanism, a dynamic evaluation model for the park response ability, and a dynamic reward and punishment response function to construct a set of efficient, secure and economic integrated energy system scheduling mechanisms.

[0032] The centralized and distributed integrated demand response method for the integrated energy system proposed by the present invention, as Figure 1 shown, includes: Step 1: Obtain the planned electricity consumption of each park, generate noise perturbations based on the differential privacy mechanism, and determine the response amount of each park using the peak shaving demand of the power grid and the planned electricity consumption after adding noise perturbations; use the difference between the planned electricity consumption and the response amount of each park as the electricity consumption benchmark value for each park to participate in the response.

[0033] Specifically, Step 1 includes: Step 1.1: Obtain the planned power consumption of each park, generate noise perturbation based on the differential privacy mechanism, and after adding the noise perturbation to the planned power consumption, the park reports it to the operator; Specifically, the park i During the peak shaving period t The planned power consumption is , and the park locally adds Noise perturbation to obtain the planned power consumption after adding the noise perturbation = + ; Among them, the noise perturbation is generated based on the differential privacy mechanism, and the probability distribution function of the noise perturbation is shown by the following relational expression:

[0034] In the formula, is the probability distribution function of the noise perturbation , is the noise control parameter, ; The noise perturbation obeys the Laplace distribution so that satisfies the differential privacy constraint, and the parameter controls the magnitude of the noise, thereby affecting the strength of privacy protection. A larger value will result in stronger privacy protection, but it will also increase the deviation of the perturbed data; by introducing this differential privacy mechanism, the park does not need to provide its sensitive power consumption data to the operator, and only needs to upload the data after noise perturbation. This protects the privacy of the park while not affecting the effectiveness of peak shaving calculation, thus realizing the unity of local retention of power consumption data and external collaborative calculation. This method significantly reduces the privacy risk when the park uploads data. In actual operation, the operator cannot directly access the real power consumption data of the park, so even in the process of participating in demand response multiple times, the privacy information of the park will not be leaked. Through this privacy protection mechanism, the park can participate in the demand response process without exposing its sensitive data, while ensuring the accuracy and efficiency of peak shaving calculation. Compared with the traditional centralized scheduling method, the present invention enhances the privacy protection ability of the park by minimizing data interaction, while ensuring the effectiveness and accuracy of peak shaving calculation.

[0035] Step 1.2: Determine the response amount of each park by using the planned power consumption after adding the noise perturbation of each park and the peak shaving demand of the power grid; Specifically, based on the planned power consumption after adding the noise perturbation of each park According to the peak shaving requirements of the superior power grid, the peak shaving demand of the power grid is allocated to each park as the response volume of each park, satisfying the following relationship:

[0036] In the formula, is the response volume of park i during the peak shaving period t . is the peak shaving demand of the power grid during the peak shaving period t . is the planned electricity consumption of park i after adding noise disturbance during the peak shaving period t .

[0037] Step 1.3: Use the difference between the planned electricity consumption and the response volume of each park as the electricity consumption benchmark value for each park to participate in the response; The electricity consumption benchmark value for each park to participate in the response satisfies the following relationship:

[0038] In the formula, is the electricity consumption benchmark value for park i to participate in the response during the peak shaving period t , is the planned electricity consumption of park i before adding noise disturbance during the peak shaving period t .

[0039] In the embodiment, a two-level optimization model architecture is proposed. In the optimization model, the integrated energy park operator is responsible for coordinating the energy demands of multiple industrial parks, optimizing the peak shaving response, ensuring that the park's energy use meets the peak shaving requirements of the power grid, and maximizing its own profit at the same time. The industrial parks participate in demand response, adjust their electricity consumption according to the requirements of the operator, optimize energy consumption, ensure that the benchmark value is met, and obtain rewards or avoid penalties. The superior power grid provides peak shaving requirements and purchases peak shaving electricity from the operator.

[0040] In the embodiment, the lower-layer optimization model mainly consists of industrial parks. The goal is to adjust its power consumption to a specified reference value according to the dispatching instructions of the integrated energy park operator, ensure participation in the peak shaving response of the power grid, and avoid excessive or insufficient power consumption. After receiving the peak shaving amount allocated by the operator, the park calculates the power consumption that needs to be responded based on the original power consumption plan of the park and compares it with the reference value. The park adjusts its power consumption according to the reference value provided by the operator. If it exceeds or is less than the reference value, rewards or penalties will be calculated based on the deviation of the adjustment amount. The park optimizes according to its own cost and reward mechanism to ensure the lowest energy consumption cost and meet the peak shaving requirements. The integrated energy park operator needs to obtain the planned power consumption of each park during the peak shaving period to allocate the peak shaving demand of the power grid and determine the response amount of the park accordingly. However, directly obtaining the original data of the park risks revealing the park's privacy. To reduce the park's provision of original energy consumption data to the integrated energy park operator, protect the park's privacy information, and balance the accuracy and efficiency of the peak shaving response calculation, in the data interaction process of the present invention, when the park reports its original power consumption plan, a data perturbation mechanism based on differential privacy is introduced to achieve data sensitivity protection.

[0041] Step 2: Determine the reward and punishment costs obtained by each park from the integrated energy park operator according to the difference between the power consumption reference value for which each park participates in the response and the actual power consumption of each park; determine the response amount corresponding to the minimum value of the difference between the operating cost of each park and the reward and punishment costs as the optimal response amount of each park.

[0042] Each park adjusts its power consumption according to the reference value and obtains corresponding rewards or penalties according to whether it exceeds or is less than the reference value. This optimization reward and punishment mechanism based on the actual response situation is different from the fixed incentive method. By dynamically adjusting the reward and punishment rules, the optimal response amount of each park can be obtained through the optimization calculation of the operating cost and the reward and punishment costs, thereby achieving the optimal economy, enhancing the enthusiasm of the park to participate in the demand response, and improving the overall energy use efficiency and system operation flexibility.

[0043] The reward and punishment costs include penalty costs and reward costs; when the actual power consumption of each park is greater than the power consumption reference value, each park obtains penalty costs from the integrated energy park operator; when the actual power consumption of each park is less than the power consumption reference value, each park obtains reward costs from the integrated energy park operator; the present invention realizes that the park adjusts its power consumption according to the reference value, and the situation of exceeding or being less than the reference value will result in corresponding rewards or penalties.

[0044] By extending the perturbation mechanism to the reward and punishment calculation link, the present invention not only improves the overall privacy protection ability of the system, but also realizes a compensation mechanism for the initial perturbation error by dynamically adjusting the reward and punishment influence coefficient, thereby enhancing the fairness of the reward and punishment mechanism and the accuracy of response scheduling without sacrificing privacy. After the comprehensive energy park operator in the park designates the electricity consumption benchmark value, it is necessary to adjust its actual electricity consumption to the electricity consumption benchmark value. If the actual electricity consumption exceeds or is less than the electricity consumption benchmark value, the park will receive corresponding penalties or rewards from the comprehensive energy park operator, satisfying the following relationship:

[0045] In the formula, is the electricity consumption benchmark value for the park i during the peak shaving period t participating in the response. When holds, is the reward for the comprehensive energy park operator to the park i during the peak shaving period t When holds, is the penalty for the comprehensive energy park operator to the park i during the peak shaving period t . is the electricity consumption of the park i during the peak shaving period t . is the unit price of the reward or penalty during the peak shaving period t . is the perturbation compensation factor of the park i during the peak shaving period t . The perturbation compensation factor satisfies the following relationship:

[0046] To further compensate for the possible impact of the noise perturbation introduced in step 1 on the accuracy of the electricity consumption benchmark value for each park participating in the response, and to ensure relatively fair compensation or exemption for parks with large noise in rewards and punishments, the present invention introduces a perturbation compensation factor in the reward and punishment calculation, and the perturbation compensation factor is coupled with the noise perturbation probability distribution through a noise control parameter, realizing a joint enhancement mechanism of perturbation compensation adjustment and differential privacy, reducing the direct impact of perturbation error on the reward and punishment mechanism, and enhancing the privacy protection and fairness of the reward and punishment mechanism.

[0047] When the industrial park participates in the peak shaving demand, its operation optimization needs to consider the rewards or punishments imposed by the integrated energy park operator on top of the operation expenditure costs. Taking the minimum difference between the operation costs and the reward and punishment costs of each park as the objective function, during the operation optimization process after the industrial park participates in the peak shaving demand, the objective function needs to comprehensively consider the operation costs and the reward and punishment costs of the park, and its goal is to minimize the difference between the operation costs and the reward and punishment costs. Specifically, the optimal response quantity can be determined by solving the following relationship, where the objective function represents the key response quantity for operation optimization. The objective function satisfies the following relationship:

[0048] In the formula, is the minimum cost of the park i , is the operation cost of the park i ; Taking the response quantity corresponding to the minimum difference between the operation cost and the reward and punishment cost of the park i as the optimal response quantity of the park i .

[0049] In the embodiment, the upper-layer optimization model focuses on the integrated energy park operator, and the goal is to maximize the operator's profit; according to the original power consumption plan reported by the park and the peak shaving demand of the power grid, the operator distributes the peak shaving tasks according to the total demand and ensures that the response quantity of the park can balance the demand of the power grid; the operator needs to consider various cost and revenue factors, including: electricity sales revenue (revenue from the peak shaving electricity purchased through the power grid), heat sales revenue (revenue from the heat energy sold through the CHP system), fuel cost (fuel cost required for the CHP system to generate electricity and heat), reward and punishment expenditure (rewarding or punishing the park to ensure that the park responds to the peak shaving demand as required); the operator adjusts the reward and punishment mechanism to encourage the park to perform demand response while maximizing its own profit.

[0050] Step 3: Taking the minimum error between the optimal response quantity and the response quantity of each park and the maximum revenue of the integrated energy park operator as the optimization goal, determine the demand response quantity instruction of each park.

[0051] The optimization goal satisfies the following relationship:

[0052] In the formula, is the optimization goal, is the revenue of the integrated energy park operator; In the ideal state, generating noise perturbation based on the differential privacy mechanism only protects the data privacy security and will not cause errors in the response quantity of each park, satisfying ​

[0053] Therefore, in the optimization of centralized and distributed integrated demand response in the integrated energy system, the present invention introduces an optimization objective that the error between the optimal response amount and the response amount of each park is minimized, and this optimization objective is the main task of each park, realizing a distributed demand response. Moreover, the planned electricity consumption after adding noise perturbation calculated in step 1 is a corrected data containing privacy protection. Since the added noise perturbation is controllable and can obtain rewards and punishments, the planned electricity consumption after adding noise perturbation can be directly used for response amount allocation and reward and punishment calculation in subsequent steps, ensuring privacy without affecting system efficiency and calculation accuracy.

[0054] In the present invention, although the differential privacy mechanism is mainly directly applied in step 1 to generate noise perturbation, its role is not limited to determining the electricity consumption benchmark value for each park to participate in the response under the condition of effectively protecting privacy. Moreover, in the process of using the difference of the electricity consumption benchmark value to determine the subsequent reward and punishment mechanism, it indirectly affects the response judgment and reward and punishment results of the entire optimization process, having persistence and global influence, and ensuring the privacy security of the park in the whole process.

[0055] In addition, maximizing the revenue of the integrated energy park operator is another optimization objective, and this optimization objective is the main task of the integrated energy park operator, realizing a centralized integrated demand response based on distributed demand response, satisfying:

[0056] The electricity sales revenue of the integrated energy park operator satisfies the following relationship:

[0057] In the formula, is the electricity sales revenue of the integrated energy park operator, is the peak shaving period 's electricity sales unit price, is the electric power output by the gas turbine during the peak shaving period ; The heat sales revenue of the integrated energy park operator satisfies the following relationship:

[0058] In the formula, is the heat sales revenue of the integrated energy park operator, is the heat sales unit price; When peak shaving is required, the power grid can purchase the required peak shaving volume from the integrated energy park operator. Specifically, the peak shaving demand of the power grid is provided by the power grid and needs to be balanced by the peak shaving response volumes of multiple parks. The optimal response volume of each park is obtained based on the goal of minimizing the difference between its cost, reward, or penalty. When the park adjusts its electricity consumption, the optimization objective function generates a corresponding optimal response volume, that is, the electricity volume required for the park to participate in peak shaving. The revenue of the integrated energy park operator from selling peak shaving volume satisfies the following relationship:

[0059] In the formula, is the revenue of the integrated energy park operator from selling peak shaving volume, is the peak shaving period t unit price of selling peak shaving volume, is the peak shaving demand of the power grid during the peak shaving period t ; The integrated energy park operator has a gas-fired combined heat and power (CHP) unit, which can obtain heat and electricity by burning natural gas. Its main output equipment inside is a gas turbine and a gas boiler. According to the variable operating condition efficiency characteristics of the gas turbine and the boiler, the relationship between its output power and fuel cost is expressed in the form of a quadratic function and satisfies the following relationship:

[0060] In the formula, is the fuel cost of the gas-fired unit during the peak shaving period t , is the electric power output by the gas-fired unit during the peak shaving period t , , , are the cost coefficients corresponding to the output of unit electric power respectively, is the heat power output by the gas-fired unit during the peak shaving period t , , , are the cost coefficients corresponding to the output of unit heat power respectively; In the centralized and distributed integrated demand response, the integrated energy park operator needs to compensate the parks participating in the response. The compensation expenditure of the integrated energy park operator is the sum of the reward and punishment costs of all parks and satisfies the following relationship:

[0061] In the formula, is the compensation expenditure of the integrated energy park operator.

[0062] Part of the peak shaving demand of the superior power grid is solved by the power generation of the integrated energy park operator. Therefore, when the error between the optimal response volume and the response volume of each park is minimized and the income of the integrated energy park operator is maximized, the demand response volume instructions of each park satisfy the following relational expression:

[0063] In the formula, is the demand response volume instruction of park i during the peak shaving period t . is the planned electricity consumption of park i during the peak shaving period t . is the peak shaving demand of the power grid during the peak shaving period t . is the electric power output by the gas turbine during the peak shaving period t .

[0064] In the embodiment, through the dual optimization mechanism, the response volume of the park and the income of the operator can reach a dynamic balance; each time the park adjusts its electricity consumption according to the benchmark value given by the operator, and at the same time reports its adjusted electricity consumption situation to the operator; the operator adjusts the strategy through the reward and punishment mechanism according to the response situation of the park to ensure that the park meets the peak shaving demand of the power grid; through multiple rounds of iteration, the park and the operator gradually optimize their respective objective functions, and finally reach an equilibrium point, meeting the peak shaving demand of the power grid while ensuring the interests of all parties.

[0065] In the traditional centralized demand response mechanism, the energy optimization of all industrial parks is usually uniformly scheduled and managed by the integrated energy park operator. Although this centralized scheduling method can ensure the unified response to the global peak shaving demand, there are also some disadvantages: First, centralized scheduling may lead to a mismatch between the energy demand of local parks and the global goal. Especially when considering the personalized energy demand, geographical distribution and operating conditions of the parks, the flexibility of the parks may not be fully utilized; Second, centralized scheduling requires large-scale data aggregation and information transmission, which may lead to a reduction in the timeliness of system response and high computational complexity. Opposite to the centralized demand response is the distributed response mechanism. Under this mechanism, each park makes decisions independently according to its own energy demand and external incentives. The advantage of distributed response is that it can improve the flexibility and response speed of the park and reduce the lag of information transmission. However, there are also some disadvantages in distributed response. Especially, the decisions between parks may not be coordinated enough, resulting in difficulty in achieving the global optimal solution. In addition, insufficient coordination between parks may lead to difficulty in accurately meeting the peak shaving demand of the power grid, thus affecting the stability of the power grid.

[0066] Through the dual optimization of the lower layer and the upper layer, in the lower layer optimization model, the park operator first allocates the peak shaving amount to each park according to the peak shaving demand of the power grid and the original power consumption plan of the park. In the upper layer model, in this model, the operator not only needs to consider the income from the power grid peak shaving service, but also needs to consider the fuel cost of its own CHP (Combined Heat and Power) system, the electricity sales income, the heat sales income, and the compensation expenditure of the park, and finely regulate the electricity and heat energy demands of the park, so as to minimize the operating cost of the park while meeting the power grid peak shaving demand and improve the reward benefit. By deeply exploring the relationship between centralized and distributed scheduling, the park can participate in demand response more flexibly, avoiding problems such as insufficient coordination and complex calculation between different levels in the traditional method. In addition, the integrated energy park operator can maximize profits through this mechanism, enhance economic benefits, and thus strengthen its competitiveness in the market.

[0067] Taking the industrial park aggregate with a voltage level of 35 kV as an example, there are 5 industrial parks in this aggregate. The number of devices in the integrated energy system of each industrial park is shown in Table 1, and the specific parameters of the devices are shown in Table 2. The interruptible load parameters of each industrial park are shown in Table 3, the time-of-use electricity price is shown in Table 4, and the unit calorific value prices of natural gas and the CHP hot steam in the park are 0.349 yuan / kWh and 0.465 yuan / kWh respectively.

[0068] Table 1 Number of devices in each park

[0069] Table 2 Specific parameters of devices

[0070] Table 3 Interruptible load parameters of each park

[0071] Table 4 Time-of-use electricity price

[0072] By adopting the method proposed in the present invention, in order to protect the privacy information such as the number and parameters of equipment in the park, each industrial park will not report all the equipment parameters in the park to the upper-level integrated energy park operator. Therefore, the integrated energy park operator cannot uniformly calculate the feasible domain. Instead, each park calculates its own feasible domain range. In order to protect its own privacy, the park needs to provide its own computing power. After calculating its own feasible domain range, each park reports it to the integrated energy park operator. The integrated energy park operator then aggregates the feasible domains of each park in the aggregate. After obtaining the aggregation results, the integrated energy park operator reports to the upper-level power grid. After obtaining the total power purchase plan and feasible domain range of the park aggregate, the upper-level power grid proposes a peak reduction of 2000kW in time period "4" based on actual needs, and sends this demand to the integrated energy park operator. The integrated energy park operator performs centralized-distributed optimization based on this demand and the process of the method proposed in the present invention. After the response, the total power purchase of the industrial park aggregate is as follows. Figure 2 As shown, each park uses energy according to the optimized energy consumption plan finally calculated by each park and reports it to the integrated energy park operator.

[0073] The integrated energy park operator first sets the peak shaving compensation or penalty electricity price for users as the basic electricity price, which is 0.35 yuan / kWh, and each iteration steps 0.01 yuan / kWh. After 10 iterations, the final peak shaving compensation electricity price is 0.26 yuan / kWh. At this time, the cost of participating in the comprehensive demand response of Park 1 is 110,286 yuan, the cost of participating in the comprehensive demand response of Park 2 is 79,603 yuan, the cost of participating in the comprehensive demand response of Park 3 is 62,425 yuan, the cost of participating in the comprehensive demand response of Park 4 is 30,315 yuan, and the cost of participating in the comprehensive demand response of Park 5 is 36,042 yuan. The integrated energy park operator makes a profit of 52,961 yuan.

[0074] Park 1 is selected as an example to analyze the park's response to peak shaving demand. As part of the entire park aggregate, after participating in the peak shaving demand response required by the upper power grid using the centralized-distributed integrated demand response method, the peak shaving demand was proposed at time period 4. Therefore, the gas turbine power generation increased significantly to meet the industrial park's electrical load demand. Since the integrated energy park operator can also provide a certain peak shaving capacity in the centralized-distributed integrated demand response method, up to 1000kW. It can be seen that since the production AC load accounts for most of the park's production load, the main means of regulating energy use is battery energy storage, and the conversion between other energy sources is an auxiliary means to enhance the flexibility of energy use in the park.

[0075] The present invention also proposes a centralized distributed integrated demand response system for an integrated energy system, comprising: The park execution module is used to obtain the planned power consumption of each park, generate noise disturbance based on the differential privacy mechanism, and determine the response amount of each park by using the peak-shaving demand of the power grid and the planned power consumption after adding noise disturbance; the difference between the planned power consumption and the response amount of each park is used as the power consumption benchmark value for each park to participate in the response; The operator execution module is used to determine the reward and punishment costs of each park from the comprehensive energy park operator based on the difference between the power consumption benchmark value of each park participating in the response and the actual power consumption of each park; the response amount corresponding to the minimum difference between the operating cost of each park and the reward and punishment cost is used as the optimal response amount of each park; The integrated execution module is used to determine the demand response quantity instructions for each park with the optimization goal of minimizing the error between the optimal response quantity and the response quantity of each park and maximizing the income of the integrated energy park operator.

[0076] In the centralized distributed integrated demand response system of the integrated energy system, the integrated energy park operator allocates the peak load according to the power consumption of the park based on the peak load demand of the upper power grid and the original power consumption plan reported by the park, and determines the peak load that each park should respond to. According to the peak load that the park should respond to, the power consumption benchmark value of the park is further calculated. In terms of architectural design, not only does it realize centralized scheduling by the integrated energy park operator, but it also introduces distributed computing and adaptive coordination mechanisms between parks, which can reduce the need for the park to provide data to the outside world while meeting the peak load demand of the power grid, protect the privacy of the park, and enhance the independent optimization capabilities of the park. In addition, the distributed computing model delegates most of the computing tasks to the park, effectively reducing the computing burden of the integrated energy park operator and improving the response speed. Each park adjusts its electricity consumption according to the benchmark value, and obtains corresponding rewards or penalties based on whether it exceeds or falls short of the benchmark value. This optimized reward and punishment mechanism based on actual response is different from the fixed incentive method. By dynamically adjusting the reward and punishment rules, the optimal response of each park can be obtained through the optimized calculation of operating costs and reward and punishment costs, thereby achieving optimal economic efficiency, enhancing the enthusiasm of the park to participate in demand response, and improving the overall energy utilization efficiency and system operation flexibility.

[0077] The integrated energy park operator optimizes the goal of maximizing profits based on the park's response. Through the two-layer optimization model, the park optimizes its own costs, and the integrated energy park operator optimizes profits, ultimately achieving the satisfaction of peak-shaving demand and maximizing the interests of all parties. Moreover, in terms of optimal scheduling, the two-layer optimization model not only optimizes the demand response of each park, but also further combines the combined heat and power (CHP) system to achieve coordinated optimal scheduling of electricity and thermal energy. The optimization goal of the two-layer optimization model not only covers the park's electricity cost, but also comprehensively considers the park's reward and punishment costs, the fuel cost of the gas units, and the operator's electricity and heat sales revenue, making the operation of the entire system more economical and reasonable.

[0078] Finally, in terms of market applicability, the technical solution of this application is not only applicable to the demand response optimal scheduling in the integrated energy system, but also can be widely applied to the intelligent scheduling of industrial parks, virtual power plants and microgrids. Different from the comparative document which only targets the regulation of the regional integrated energy system, through the two-layer optimization model, this application enables the integrated energy system to more flexibly adapt to market changes, improve energy utilization efficiency, and maximize the benefits of operators and parks while ensuring the stability of the power grid.

[0079] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications or equivalent replacements can still be made to the specific implementation manners of the present invention, and any modification or equivalent replacement that does not depart from the spirit and scope of the present invention shall be covered by the protection scope of the claims of the present invention.

Claims

1. A centralized distributed integrated demand response method for an integrated energy system, characterized in that: include: Obtain the planned electricity consumption of each park, generate noise disturbance based on the differential privacy mechanism, and use the peak-shaving demand of the power grid and the planned electricity consumption after adding noise disturbance to determine the response amount of each park; the difference between the planned electricity consumption and the response amount of each park is used as the electricity consumption benchmark value for each park to participate in the response; According to the difference between the power consumption benchmark value of each park participating in the response and the actual power consumption of each park, determine the reward and punishment cost of each park from the integrated energy park operator; the response amount corresponding to the minimum difference between the operating cost of each park and the reward and punishment cost is taken as the optimal response amount of each park; The demand response quantity instructions for each park are determined with the optimization goal of minimizing the error between the optimal response quantity and the response quantity of each park and maximizing the income of the integrated energy park operator.

2. The centralized distributed integrated demand response method of the integrated energy system according to claim 1 is characterized in that: Obtain the planned electricity consumption of each park, generate noise disturbance based on the differential privacy mechanism, and report the noise disturbance to the planned electricity consumption to the operator after adding it. Park i During peak hours t The planned electricity consumption is , the park is locally Add noise disturbance , get the planned power consumption after adding noise disturbance = + ; Among them, noise perturbation is generated based on differential privacy mechanism. The probability distribution function of is shown in the following relationship: In the formula, Noise disturbance The probability distribution function of is the noise control parameter, .

3. The centralized distributed integrated demand response method of the integrated energy system according to claim 2 is characterized in that: Planned electricity consumption after adding noise disturbance in each park and the peak-shaving demand of the upper power grid, and distribute the peak-shaving demand of the power grid to each park as the response amount of each park, satisfying the following relationship: In the formula, For the park i During peak hours t The response amount, For the power grid during peak load period t The peak demand, For the park i During peak hours t Increase the planned electricity consumption after noise disturbance.

4. The centralized distributed integrated demand response method for an integrated energy system according to claim 3, characterized in that: The difference between the planned power consumption of each park and the response amount is used as the power consumption benchmark value for each park to participate in the response, satisfying the following relationship: In the formula, For the park i During peak hours t The power consumption benchmark value for participating responses, For the park i During peak hours t Planned electricity consumption.

5. The centralized and distributed integrated demand response method of the integrated energy system according to claim 4 is characterized in that: The park receives corresponding punishment or reward from the integrated energy park operator, satisfying the following relationship: In the formula, when hour, Provide comprehensive energy park operators with i During peak hours t The reward, when hour, Provide comprehensive energy park operators with i During peak hours t punishment, For the park i During peak hours t of electricity consumption, Peak load period t The unit price of reward or punishment, For the park i During peak hours t The disturbance compensation factor.

6. The centralized distributed integrated demand response method of the integrated energy system according to claim 5, characterized in that: The disturbance compensation factor satisfies the following relationship: 。 7. The centralized distributed integrated demand response method for an integrated energy system according to claim 1, characterized in that: The optimization objective satisfies the following relationship: In the formula, To optimize the goal, For the park i During peak hours t The optimal response quantity, For the park i During peak hours t The response amount, Revenue for integrated energy park operators.

8. The centralized and distributed integrated demand response method of the integrated energy system according to claim 7, characterized in that: In the formula, The electricity sales revenue for the integrated energy park operators, Peak load period t The unit price of electricity sales, For gas units during peak load period t Output electrical power; In the formula, Heat sales revenue for integrated energy park operators, The unit price for hot sale; In the formula, Revenue from selling peak load to integrated energy park operators, Peak load period t The unit price of selling peak load, For the power grid during peak load period t Peak load demand; In the formula, For gas units during peak load period t The fuel cost, For gas units during peak load period t The output electrical power, , , are the cost coefficients corresponding to the output unit electric power, For gas units during peak load period t Output heat power, , , are the cost coefficients corresponding to the output unit thermal power respectively; In the formula, Compensation expenses for integrated energy park operators, Provide comprehensive energy park operators with i During peak hours t rewards or punishments.

9. The centralized distributed integrated demand response method of the integrated energy system according to claim 8, characterized in that: When the error between the optimal response amount and the response amount of each park is the smallest and the income of the integrated energy park operator is the largest, the demand response amount instructions of each park satisfy the following relationship: In the formula, For the park i During peak hours t Demand response volume instructions, For the park i During peak hours t Planned electricity consumption, For the power grid during peak load period t The peak demand, For gas units during peak load period t Output electrical power.

10. A centralized distributed integrated demand response system for an integrated energy system, characterized in that: include: The park execution module is used to obtain the planned power consumption of each park, generate noise disturbance based on the differential privacy mechanism, and determine the response amount of each park by using the peak-shaving demand of the power grid and the planned power consumption after adding noise disturbance; the difference between the planned power consumption and the response amount of each park is used as the power consumption benchmark value for each park to participate in the response; The operator execution module is used to determine the reward and punishment costs of each park from the comprehensive energy park operator based on the difference between the power consumption benchmark value of each park participating in the response and the actual power consumption of each park; the response amount corresponding to the minimum difference between the operating cost of each park and the reward and punishment cost is used as the optimal response amount of each park; The integrated execution module is used to determine the demand response quantity instructions of each park with the optimization goal of minimizing the error between the optimal response quantity and the response quantity of each park and maximizing the income of the integrated energy park operator.

Images