An industrial load multi-energy flexibility evaluation method considering uncertainty

Abstract

This invention discloses a method for assessing the multi-energy flexibility of industrial loads considering uncertainties. The method includes: constructing a model of the energy consumption and output relationship of a production process within an industrial load considering output adjustment constraints; using graph theory to associate the production process and product storage warehouses to construct storage and output requirement constraints; constructing a discrete set of states for the industrial load and performing a feasible interval search to construct a multi-energy feasible interval; searching the feasible interval to obtain the multi-energy flexibility of the industrial load, thus achieving the assessment of multi-energy flexibility. This invention assesses the multi-energy flexible adjustment capability of industrial loads considering production uncertainties. It can be used to adjust the demand of industrial loads for various energy forms at the operational level, and can be applied to multiple fields such as electricity and multi-energy demand response. This assists energy systems in utilizing the adjustment capability of industrial loads, enhancing system flexibility, and supporting the flexible operation and optimized scheduling of energy systems.

Description

Technical Field

[0001] This invention relates to a multi-energy flexibility assessment method, specifically a multi-energy flexibility assessment method for industrial loads that takes uncertainty into account. Background Technology

[0002] Industrial loads are a significant component of energy systems, representing the largest share of energy consumption. With the development of energy systems and the increasing proportion of renewable energy sources, greater flexibility is needed for absorption and balancing. Therefore, at the system operation level, the adjustability of industrial loads can be rationally utilized to provide flexibility. However, industrial loads themselves are inherently uncertain, exhibiting fluctuations in output and changes in orders. The flexible resources required by the energy system must be deterministic and controllable; otherwise, it would increase the burden of flexible scheduling. Currently, there is a lack of high-reliability research that considers the uncertainties of the multi-energy flexibility of industrial loads in constructing stable and reliable methods for flexible interaction within the energy system. Summary of the Invention

[0003] This invention proposes a method for assessing the multi-energy flexibility of industrial loads that takes into account uncertainties. This method can obtain the multi-energy adjustable range of industrial loads with probabilistic values. This adjustable capability can be used by the energy system for optimal scheduling, thereby enhancing the operational flexibility of the energy system.

[0004] The technical solution adopted in this invention is:

[0005] The industrial load multi-energy flexibility assessment method of the present invention includes the following steps:

[0006] 1) Construct a model of the relationship between energy consumption and output in an industrial load considering output adjustment constraints.

[0007] 2) Use graph theory to associate the production process and product storage warehouse in the industrial load in step 1), and construct the storage constraints of the product storage warehouse and the output requirement constraints of the production process based on the serial and parallel relationships between the various production processes.

[0008] 3) Treat the uncertainties in the production process of industrial load as random variables, discretize the random variables to construct a set of discrete states of industrial load. Each set of discrete states includes several discrete states of industrial load. Based on the energy consumption and output relationship model, search the feasible interval for output adjustment constraints, product storage warehouse storage constraints and production process output requirement constraints under each discrete state to construct the multi-energy feasible interval and their respective probabilities under each discrete state.

[0009] 4) Based on the preset probability value requirement of the multi-energy flexibility of the industrial load, search for feasible intervals that meet the preset probability value requirement in the multi-energy feasible intervals under each discrete state, obtain the multi-energy flexibility of the industrial load, and thus realize the evaluation of the multi-energy flexibility of the industrial load.

[0010] The uncertainty referred to here refers to the uncertainty in the industrial production process. Uncertainty in the industrial production process includes factors such as equipment failure, temporary arrival or cancellation of orders, changes in output and inventory caused by fluctuations in unit output, and changes in output caused by changes in equipment parameters. These uncertainties affect the energy flow and material flow balance of the system. To characterize these uncertainties, random variables can be used in subsequent modeling. The distribution of these random variables can be an empirical distribution obtained from historical data, or a commonly used probability distribution model such as the normal distribution or the Poisson distribution can be used.

[0011] The production process in the industrial load refers to a link in industrial production, distinguished by workshops or processes. The criterion for this division is that there are no obvious time constraints between different production processes; that is, materials or products processed in the previous production process can be stored without immediately entering the next. Dividing industrial loads into production processes is beneficial for industrial production management. Different production processes produce stable intermediate products or materials, facilitating scheduling and manpower allocation for different processes. From an equipment perspective, a production process may involve only one type of equipment or a processing system composed of multiple devices. When the production process is a processing system composed of multiple devices, it indicates strict time and sequence constraints on materials during the process; that is, after processing by one device, materials need to enter the next device within a short period. The processing order of materials between these devices is also strictly required. After the entire processing stage, materials can be stored. Many industries already have production processes divided. Based on this, to explore the flexibility of different production processes in industrial loads, they are divided into three categories according to their energy consumption-output characteristics: discrete production processes, continuous production processes, and flexible production processes. An energy consumption and output relationship model is established for each category.

[0012] In step 1), the production processes in industrial loads are classified according to their production characteristics. These include discrete production processes, continuous production processes, and flexible production processes. The energy consumption and output relationship model for production processes in industrial loads is as follows:

[0013] The energy consumption and output relationship model for discrete manufacturing processes is as follows:

[0014]

[0015]

[0016] in, Indicates the output of a discrete manufacturing process; This represents the number of production lines that have been started in the discrete manufacturing process. Since the number of production lines in the discrete manufacturing process is relatively small, the uncertainty of the number of start-up and shutdown is also relatively small. This value is a decision variable. This represents the output of a single production line within a discrete production process, and is a known quantity. This represents the total number of production lines in a discrete production process, and is a known quantity. This represents a column vector of various energy consumptions in a discrete production process, including the consumption of electrical energy, natural gas energy, heat energy, etc. This represents a column vector of various energy consumptions in a single production line within a discrete production process. The various energy consumption column vectors include the consumption of electrical energy, natural gas energy, heat energy, etc., which are known quantities.

[0017] The discrete production process refers to a process where energy consumption and output switch only between a few fixed points. Equipment within a discrete production process is often characterized by high individual power, small quantity, and either lack of individual adjustability or high difficulty in control. The regulation and control of individual equipment or a single production process is a key research focus in chemical production process control, aiming to achieve control on a short timescale to stabilize various indicators in the production process. Therefore, regarding the hourly-level adjustment capability focused on by this patent, production has already reached a steady state, which is a fixed operating point of the discrete production process. Typical discrete production processes are most processes in chemical production, such as smelting and electrolysis. When production needs adjustment, such processes can only shut down or start part of the production line, resulting in energy consumption and output switching only between a few operating points.

[0018] The energy consumption and output relationship model for a continuous production process is as follows:

[0019]

[0020] in, This represents a column vector of various energy consumptions in a continuous production process, including the consumption of electrical energy, natural gas energy, heat energy, etc. Indicates the output of a continuous production process; The relationship between various energy consumption and output in a continuous production process can be obtained from the unit output and rated energy consumption of a single production line within the continuous production process, and is a known quantity.

[0021] The continuous production process refers to a process where energy consumption and output can be continuously adjusted over a large range. Individual production lines within a continuous production process are characterized by low power, large quantity, and non-adjustability. Compared to production lines in discrete production processes, individual production lines in a continuous production process have lower power, so multiple identical production lines are often used for simultaneous production. In actual production, the number of production lines is typically between 20 and 60, which can be considered a continuous production process. Typical continuous production processes are weaving and injection molding processes; a textile mill often has 80-150 weaving machines. This type of production process also adjusts output and energy consumption by shutting down or starting some production lines. However, due to its low individual power and large total output, it is one of the main adjustable resources; therefore, when assessing the adjustment capability of industrial load, energy consumption and output are approximately considered to be continuously adjustable.

[0022] The energy consumption and output relationship model for flexible production processes is as follows:

[0023]

[0024] in, This represents a column vector of various energy consumptions in a flexible production process, including the consumption of electrical energy, natural gas energy, heat energy, etc. Indicates output from a flexible production process; The linear terms representing the relationship between energy consumption and output in flexible production processes are known quantities. The constant term represents the relationship between various energy consumption and output in flexible production processes, and is a known quantity.

[0025] The specific production adjustment constraints are as follows:

[0026]

[0027]

[0028] in, and Let represent the minimum and maximum output of the discrete production process within a scheduling step, respectively. Considering the uncertainties in the production process, these two values ​​are represented by random variables with known distributions. and Let represent the minimum and maximum output of the flexible production process within a scheduling step, respectively. Considering the uncertainties in the production process, these two values ​​are represented by random variables with known distributions.

[0029] The flexible production process refers to a process where energy consumption and output can be continuously adjusted within a small range. A flexible production process typically involves only one or a few production lines, but these lines are highly intelligent and automated, characterized by repetitive production and the ability to produce a considerable number of products in a short time, usually on the order of tens of units per minute. Therefore, considering a flexible scheduling step size of one hour, continuous output control can be approximately achieved. However, unlike continuous production processes, the adjustment principle of a flexible production process is based on adjusting the entire production line rather than starting or stopping parts of it. Therefore, the adjustment capability is limited by the physical constraints of the production line, and the adjustment range is smaller compared to continuous production processes. A typical flexible production process is the surface mount technology (SMT) production process, where the chip placement speed is often tens or even hundreds of pieces per minute, and this speed can be adjusted through settings to change the energy consumption of the equipment.

[0030] In step 2), graph theory is used to connect the production processes and product storage warehouses in the industrial load of step 1). Specifically, the production processes and product storage warehouses are modeled as nodes, and the material flow between them is represented as the connection between nodes. Because the direction of material flow is definite, the entire industrial load can be connected based on the relationships between each production process and product storage warehouse, constructing its topology. Once the topology is clear, mathematical symbols can be used to explicitly represent the connections between each production process, what the input materials required for a particular production process are, and what its output product is. Output requirement constraints are based on the series and parallel relationships between the various production processes.

[0031] In step 2), the storage constraints of the product storage warehouse are as follows:

[0032]

[0033] in, and These represent the maximum material outflow and material inflow within a scheduling step of the production process, respectively, and are known quantities. and These represent the minimum and maximum material storage quantities in the product storage warehouse, respectively, and are known quantities. This represents the change in the inventory of the product storage warehouse within a scheduling step, and is a decision variable; This represents the inventory of the product storage warehouse at time t-1 before the scheduling period at time t, and is a known quantity.

[0034] In step 2), the specific output requirements of the production process are as follows:

[0035]

[0036] in, A column vector representing the change in inventory of all product storage warehouses within the industrial load; This is a column vector representing the existing inventory of all product storage warehouses within the industrial load at time t-1. The time of the entire scheduling cycle of the production process is a known quantity; A column vector representing the maximum output of all production processes; A relational matrix representing the production process and product storage warehouses; This represents a column vector of target outputs for each production process within a scheduling cycle, and these are known quantities.

[0037] The correlation matrix between the production process and the product storage warehouse. It includes several related elements, as follows:

[0038]

[0039] in, An association matrix representing the production process and product storage warehouses. The associated element in the i-th row and j-th column of the array; Indicates the product conversion factor; This represents the nodes in the industrial load topology associated by graph theory. To the node The product of the raw material and product conversion coefficients for all production processes is a known quantity; the output node is specifically the last node in the industrial load topology associated by graph theory.

[0040] The aforementioned output requirement constraint based on the series-parallel relationships between various production processes refers to the fact that industrial load provides flexibility by changing the output of certain production processes within a scheduling step. Therefore, in order to meet its production tasks throughout the entire scheduling cycle, the industrial load needs to consider its ability to compensate for the adjusted output in subsequent time periods. Accordingly, based on the premise that flexibility is invoked at most once per scheduling cycle, output requirement constraints for each production link are constructed.

[0041] In step 3), under the constraints of storage in the product storage warehouse and output requirements in the production process, the random variables are discretized to construct a set of discrete states for the industrial load. For each set of discrete states... The details are as follows:

[0042]

[0043] in, and These represent the minimum and maximum output of the discrete production process within one scheduling step when the random variable is in its first discrete state; and These represent the minimum and maximum output of the flexible production process within a scheduling step when the random variable is in its first discrete state; and These represent the minimum and maximum output of the discrete production process within one scheduling step when the random variable is in its second discrete state; and These represent the minimum and maximum output of the flexible production process within one scheduling step when the random variable is in its second discrete state; and Each of the random variables in the industrial load represents the state at the _th ... For a discrete state, the minimum and maximum output of a discrete production process within a scheduling step; and Each of the random variables in the industrial load represents the state at the _th ... For a discrete state, the minimum and maximum output of a flexible production process within a scheduling step; and Each of the random variables in the industrial load represents the state at the _th ... For a discrete state, the minimum and maximum output of a discrete production process within a scheduling step; and Each of the random variables in the industrial load represents the state at the _th ... For a set of discrete states, the minimum and maximum output of the flexible production process within a scheduling step; the discrete state set This represents the state of each random variable in the industrial load. At that time, the state of the entire industrial load.

[0044] Taking the maximum unit output of a continuous production process as an example, the process of discretizing random variables is divided into... There are several states, and the number of states represents the parameters that can be set and adjusted. The entire industrial load contains... The discrete state of industrial load has several random variables. There are several discrete states, and the probability of each discrete state can be obtained by multiplying the states of these random variables together.

[0045] The random variable is at the th The probability of a discrete state can be obtained from the probability density function of the random variable, as follows:

[0046]

[0047] in, Indicates that the random variable is in the th position. The probability of a discrete state; The density function of a random variable, as mentioned earlier, can be an empirical distribution obtained from historical data, or it can use a commonly used probability distribution model such as the normal distribution or the Poisson distribution, and is a known quantity.

[0048] In step 3), based on the energy consumption and output relationship model, a feasible interval search is performed on the output adjustment constraint, the storage constraint of the product storage warehouse, and the output requirement constraint of the production process under each discrete state to construct a multi-energy feasible interval for each discrete state. Specifically, based on the output adjustment constraint in step 1) and the storage constraint of the product storage warehouse and the output requirement constraint of the production process in step 2), the vertex enumeration method is used to search for the feasible set of all random variables. After obtaining the feasible set of all random variables under each discrete state, the output and energy consumption relationship model of the production process in step 1) outputs the multi-energy feasible interval of the industrial load under each discrete state.

[0049] The vertex enumeration method specifically involves first denoting the number of discrete production processes in the industrial load as follows: The number of continuous production processes is denoted as The number of flexible production processes is denoted as The number of product storage warehouses is denoted as The independent variables in the production and energy consumption relationship model are: This is because the variables representing the output of the production process and the variables representing the increment of the product storage warehouse are independent of each other, and the system only has these independent variables. Then, the feasible interval is searched by traversing the output adjustment constraints in the industrial load. Since the variables of the discrete industrial process are integer variables, while the variables of the continuous production process and the flexible production process are continuous variables, the impact on the flexibility of the industrial load will be different, as follows:

[0050] First fix The values ​​of the discrete production process variables are randomly selected from the output adjustment constraints in step 1), the storage constraints of the product storage warehouse in step 2), and the output requirement constraints of the production process. Constraints, determine by If the judgment matrix composed of the criteria is full rank, then a feasible solution for a set of independent variables is obtained, thus determining the number of production lines already started in each discrete production process. The change in inventory of the product storage warehouse within a scheduling step. A set of solutions represents the vertices of these variables in the decision space; if not, the process continues until a solution is found. If the judgment matrix composed of the criteria is full rank, a feasible solution for a set of independent variables is obtained; therefore, this method searches all vertices in the variable space to determine the adjustable domain of the decision variables.

[0051] After completing the traversal, several feasible solutions with independent variables are obtained, which means the number of production lines that have been started in each discrete production process. Output of continuous production processes Output from flexible production processes The change in inventory of the product storage warehouse within a scheduling step. Given several sets of values ​​for the independent variables, the feasible solutions of each set of independent variables constitute the solution set of the multi-energy demand feasible solution. The feasible solutions of each set of independent variables are denoted as a matrix. subscript This indicates that the solution set is the solution set under a certain system state.

[0052] The feasible solutions for each set of independent variables in the feasible solution set of multi-energy demand are transformed into feasible solutions for various energy consumptions through a production process output and energy consumption relationship model, as follows:

[0053]

[0054] in, Indicates that the random variable is in the th position. The set of feasible solutions for multi-energy demand under discrete states; The transformation matrix representing the control variables of output and energy consumption in the production process can be obtained from the material and energy relationships in steps 1) and 2).

[0055] For all solutions in the feasible solution set of multi-energy demand, find the maximum and minimum values ​​of a single energy source to obtain the next... A multi-functional feasible interval for industrial load under discrete conditions.

[0056] In step 4), based on the preset probability value requirement for the multi-energy flexibility of the industrial load, a feasible interval that meets the preset probability value requirement is searched in the multi-energy feasible intervals under each discrete state to obtain the multi-energy flexibility of the industrial load, as detailed below:

[0057] In the entire industrial load There are n random variables, each random variable includes There are discrete states for the industrial load. One, according to step 3), obtain the industrial load. For each discrete state of the multi-energy feasible interval, the flexible adjustment interval of each energy is searched. For each energy, the maximum and minimum values ​​of the feasible interval of the energy in all discrete states are first obtained. Starting from the minimum value, the search calculation begins with a preset equal step size. The probability of each step size point is calculated, that is, the probability values ​​of the step size points in all discrete states are added together. Finally, after obtaining the probability values ​​of each step size point in the flexible adjustment interval of the energy, the flexible adjustment interval of the energy that meets the preset probability value requirements is selected, and finally the multi-energy flexibility of the industrial load is obtained.

[0058] The beneficial effects of this invention are:

[0059] This invention provides a highly reliable assessment method for industrial loads to participate in flexible interaction with energy systems. The method takes into account the regulation characteristics of different production links in industrial loads and the possible uncertainties when they participate in regulation. Finally, it uses probability values ​​to reflect the impact of uncertainty on regulation capacity in the multi-energy regulation range of industrial loads, thereby clarifying the relationship between regulation capacity and uncertainty, providing a reliable assessment of the regulation capacity of energy systems, and assisting energy systems in formulating various flexible operation plans.

[0060] The method of this invention evaluates the multi-energy flexible adjustment capability of industrial loads considering production uncertainties. It can be used to adjust the demand of industrial loads for various energy forms at the operational level. It can be applied to multiple fields such as electricity and multi-energy demand response, thereby assisting energy systems in utilizing the adjustment capability of industrial loads, enhancing system flexibility, and supporting the flexible operation and optimized scheduling of energy systems. Attached Figure Description

[0061] Figure 1 This is a flowchart of the evaluation method of the present invention;

[0062] Figure 2 This is a schematic diagram illustrating the connection between the production process and the product storage warehouse based on graph theory in an embodiment of the present invention.

[0063] Figure 3 This is a feasible range of industrial load under a system state according to an embodiment of the present invention;

[0064] Figure 4 The embodiment of the present invention considers the uncertainty of the industrial load and has probability value requirements, which is a flexible and feasible range. Detailed Implementation

[0065] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0066] like Figure 1 As shown, the industrial load multi-energy flexibility assessment method of the present invention includes the following steps:

[0067] 1) Construct a model of the relationship between energy consumption and output in an industrial load considering output adjustment constraints.

[0068] Uncertainty refers to the uncertainty in industrial production processes. Uncertainty in industrial production processes includes factors such as equipment failures, unexpected order arrivals or cancellations, fluctuations in unit output leading to changes in output and inventory, and changes in equipment parameters causing changes in output. These uncertainties affect the energy and material flow balance of the system. To characterize these uncertainties, random variables can be used in subsequent modeling. The distribution of these random variables can be an empirical distribution derived from historical data, or a commonly used probability distribution model such as the normal distribution or the Poisson distribution can be used.

[0069] In industrial loads, a production process refers to a stage in industrial production, distinguished by workshops or processes. The criterion for this division is the absence of clear time constraints between different production processes; that is, materials or products processed in one process can be stored without immediately entering the next. Dividing industrial loads into production processes facilitates production management. Different processes produce stable intermediate products or materials, making it easier for production management to schedule shifts and allocate manpower. From an equipment perspective, a production process may involve only one type of equipment or a processing system composed of multiple devices. When a production process is a processing system composed of multiple devices, it indicates strict time and sequence constraints on materials during the process; that is, after processing on one device, materials need to enter the next device within a short period. The processing order of materials between these devices is also strictly required. After the entire processing stage, materials can be stored. Many industries already have production processes divided. Based on this, to explore the flexibility of different production processes in industrial loads, they are divided into three categories according to their energy consumption-output characteristics: discrete production processes, continuous production processes, and flexible production processes. An energy consumption and output relationship model is established for each category.

[0070] In step 1), the production processes in the industrial load are classified according to their production characteristics. The production processes in the industrial load include discrete production processes, continuous production processes, and flexible production processes. The energy consumption and output relationship model of the production processes in the industrial load is as follows:

[0071] The energy consumption and output relationship model for discrete manufacturing processes is as follows:

[0072]

[0073]

[0074] in, Indicates the output of a discrete manufacturing process; This represents the number of production lines that have been started in the discrete manufacturing process. Since the number of production lines in the discrete manufacturing process is relatively small, the uncertainty of the number of start-up and shutdown is also relatively small. This value is a decision variable. This represents the output of a single production line within a discrete production process, and is a known quantity. This represents the total number of production lines in a discrete production process, and is a known quantity. This represents a column vector of various energy consumptions in a discrete production process, including the consumption of electrical energy, natural gas energy, heat energy, etc. This represents a column vector of various energy consumptions in a single production line within a discrete production process. The various energy consumption column vectors include the consumption of electrical energy, natural gas energy, heat energy, etc., which are known quantities.

[0075] Discrete production processes refer to processes where energy consumption and output switch only between a few fixed points. Equipment within discrete production processes is often characterized by high individual power, small quantity, and either lack of individual adjustability or high difficulty in control. The regulation and control of individual equipment or a single production process is a key research focus in chemical production process control, aiming to achieve control on a short timescale to stabilize various indicators in the production process. Therefore, regarding the hourly-level adjustment capability focused on by this patent, production has already reached a steady state, which is a fixed operating point of the discrete production process. Typical discrete production processes are most processes in chemical production, such as smelting and electrolysis. When production needs adjustment, such processes can only shut down or start part of the production line, resulting in energy consumption and output switching only between a few operating points.

[0076] The energy consumption and output relationship model for a continuous production process is as follows:

[0077]

[0078] in, This represents a column vector of various energy consumptions in a continuous production process, including the consumption of electrical energy, natural gas energy, heat energy, etc. Indicates the output of a continuous production process; The relationship between various energy consumption and output in a continuous production process can be obtained from the unit output and rated energy consumption of a single production line within the continuous production process, and is a known quantity.

[0079] A continuous production process refers to a process where energy consumption and output can be adjusted approximately continuously over a large range. Individual production lines within a continuous production process are characterized by low power, large quantity, and non-adjustable characteristics. Compared to production lines in discrete production processes, individual production lines in a continuous process have lower power consumption, so multiple identical production lines are often used for simultaneous production. In actual production, the number of production lines is typically between 20 and 60, which can be considered a continuous production process. Typical continuous production processes include weaving and injection molding, where a textile mill often has 80-150 weaving machines. This type of production process also adjusts output and energy consumption by shutting down or starting some production lines. However, due to its low individual power consumption and large total output, it is one of the main adjustable resources; therefore, when assessing the adjustment capability of industrial load, energy consumption and output are approximated as continuously adjustable.

[0080] The energy consumption and output relationship model for flexible production processes is as follows:

[0081]

[0082] in, This represents a column vector of various energy consumptions in a flexible production process, including the consumption of electrical energy, natural gas energy, heat energy, etc. Indicates output from a flexible production process; The linear terms representing the relationship between energy consumption and output in flexible production processes are known quantities. The constant term represents the relationship between various energy consumption and output in flexible production processes, and is a known quantity.

[0083] The specific constraints on production adjustment are as follows:

[0084]

[0085]

[0086] in, and Let represent the minimum and maximum output of the discrete production process within a scheduling step, respectively. Considering the uncertainties in the production process, these two values ​​are represented by random variables with known distributions. and Let represent the minimum and maximum output of the flexible production process within a scheduling step, respectively. Considering the uncertainties in the production process, these two values ​​are represented by random variables with known distributions.

[0087] Flexible production processes refer to processes where energy consumption and output can be continuously adjusted within a small range. Flexible production processes typically involve only one or a few production lines, but these lines are highly intelligent and automated, characterized by repetitive production that can produce a considerable number of products in a short time, usually on the order of tens of units per minute. Therefore, considering a flexible scheduling step size of one hour, continuous output control can be approximately achieved. However, unlike continuous production processes, the adjustment principle of flexible production processes is based on adjusting the entire production line rather than starting or stopping parts of it. Therefore, the adjustment capability is limited by the physical constraints of the production line, and the adjustment range is smaller compared to continuous production processes. A typical flexible production process is the surface mount technology (SMT) production process, where the chip placement speed is often tens or even hundreds of units per minute, and this speed can be adjusted through settings to change the energy consumption of the equipment.

[0088] 2) Use graph theory to associate the production process and product storage warehouse in the industrial load in step 1), and construct the storage constraints of the product storage warehouse and the output requirement constraints of the production process based on the serial and parallel relationships between the various production processes.

[0089] In step 2), graph theory is used to connect the production processes and product storage warehouses in the industrial load of step 1). Specifically, the production processes and product storage warehouses are modeled as nodes, and the material flow between them is represented as the connection between nodes. Because the direction of material flow is definite, the entire industrial load can be connected based on the relationships between each production process and product storage warehouse, constructing its topology. Once the topology is clear, mathematical symbols can be used to explicitly represent the connections between each production process, what the input materials required for a particular production process are, and what its output product is. Output requirement constraints are then established based on the series and parallel relationships between the various production processes.

[0090] In step 2), the storage constraints of the product storage warehouse are as follows:

[0091]

[0092] in, and These represent the maximum material outflow and material inflow within a scheduling step of the production process, respectively, and are known quantities. and These represent the minimum and maximum material storage quantities in the product storage warehouse, respectively, and are known quantities. This represents the change in the inventory of the product storage warehouse within a scheduling step, and is a decision variable; This represents the inventory of the product storage warehouse at time t-1 before the scheduling period at time t, and is a known quantity.

[0093] In step 2), the specific output requirements constraints for the production process are as follows:

[0094]

[0095] in, A column vector representing the change in inventory of all product storage warehouses within the industrial load; This is a column vector representing the existing inventory of all product storage warehouses within the industrial load at time t-1. The time of the entire scheduling cycle of the production process is a known quantity; A column vector representing the maximum output of all production processes; A relational matrix representing the production process and product storage warehouses; This represents a column vector of target outputs for each production process within a scheduling cycle, and these are known quantities.

[0096] Relationship matrix between production process and product storage warehouse It includes several related elements, as follows:

[0097]

[0098] in, An association matrix representing the production process and product storage warehouses. The associated element in the i-th row and j-th column of the array; Indicates the product conversion factor; This represents the nodes in the industrial load topology associated by graph theory. To the node The product of the raw material and product conversion coefficients for all production processes is a known quantity; the output node is specifically the last node in the industrial load topology associated by graph theory.

[0099] Output requirement constraints based on the series-parallel relationships between various production processes refer to the situation where industrial loads provide flexibility by altering the output of certain production processes within a scheduling step. Therefore, throughout the entire scheduling cycle, in order to meet its production tasks, the industrial load needs to consider its ability to compensate for the adjusted output in subsequent time periods. Based on this, and assuming that flexibility is invoked at most once per scheduling cycle, output requirement constraints for each production stage are constructed.

[0100] 3) Treat the uncertainties in the production process of industrial load as random variables, discretize the random variables to construct a set of discrete states of industrial load. Each set of discrete states includes several discrete states of industrial load. Based on the energy consumption and output relationship model, search the feasible interval for output adjustment constraints, product storage warehouse storage constraints and production process output requirement constraints under each discrete state to construct the multi-energy feasible interval and their respective probabilities under each discrete state.

[0101] In step 3), under the constraints of storage in the product storage warehouse and output requirements in the production process, the random variables are discretized to construct a set of discrete states for the industrial load. For each set of discrete states... The details are as follows:

[0102]

[0103] in, and These represent the minimum and maximum output of the discrete production process within one scheduling step when the random variable is in its first discrete state; and These represent the minimum and maximum output of the flexible production process within a scheduling step when the random variable is in its first discrete state; and These represent the minimum and maximum output of the discrete production process within one scheduling step when the random variable is in its second discrete state; and These represent the minimum and maximum output of the flexible production process within one scheduling step when the random variable is in its second discrete state; and Each of the random variables in the industrial load represents the state at the _th ... For a discrete state, the minimum and maximum output of a discrete production process within a scheduling step; and Each of the random variables in the industrial load represents the state at the _th ... For a discrete state, the minimum and maximum output of a flexible production process within a scheduling step; and Each of the random variables in the industrial load represents the state at the _th ... For a discrete state, the minimum and maximum output of a discrete production process within a scheduling step; and Each of the random variables in the industrial load represents the state at the _th ... For a set of discrete states, the minimum and maximum output of the flexible production process within a scheduling step; the discrete state set This represents the state of each random variable in the industrial load. At that time, the state of the entire industrial load.

[0104] Taking the maximum unit output of a continuous production process as an example, the process of discretizing random variables is divided into... There are several states, and the number of states represents the parameters that can be set and adjusted. The entire industrial load contains... The discrete state of industrial load has several random variables. There are several discrete states, and the probability of each discrete state can be obtained by multiplying the states of these random variables together.

[0105] The random variable is at the th The probability of a discrete state can be obtained from the probability density function of the random variable, as follows:

[0106]

[0107] in, Indicates that the random variable is in the th position. The probability of a discrete state; The density function of a random variable, as mentioned earlier, can be an empirical distribution obtained from historical data, or it can use a commonly used probability distribution model such as the normal distribution or the Poisson distribution, and is a known quantity.

[0108] In step 3), based on the energy consumption and output relationship model, a feasible interval search is performed on the output adjustment constraint, the storage constraint of the product storage warehouse, and the output requirement constraint of the production process under each discrete state to construct the multi-energy feasible interval for each discrete state. Specifically, based on the output adjustment constraint in step 1) and the storage constraint of the product storage warehouse and the output requirement constraint of the production process in step 2), the vertex enumeration method is used to search for the feasible set of all random variables. After obtaining the feasible set of all random variables under each discrete state, the output and energy consumption relationship model of the production process in step 1) outputs the multi-energy feasible interval of the industrial load under each discrete state.

[0109] The vertex enumeration method specifically involves first denoting the number of discrete production processes in the industrial load as follows: The number of continuous production processes is denoted as The number of flexible production processes is denoted as The number of product storage warehouses is denoted as The independent variables in the production and energy consumption relationship model are: This is because the variables representing the output of the production process and the variables representing the increment of the product storage warehouse are independent of each other, and the system only has these independent variables. Then, the feasible interval is searched by traversing the output adjustment constraints in the industrial load. Since the variables of the discrete industrial process are integer variables, while the variables of the continuous production process and the flexible production process are continuous variables, the impact on the flexibility of the industrial load will be different, as follows:

[0110] First fix The values ​​of the discrete production process variables are randomly selected from the output adjustment constraints in step 1), the storage constraints of the product storage warehouse in step 2), and the output requirement constraints of the production process. Constraints, determine by If the judgment matrix composed of the criteria is full rank, then a feasible solution for a set of independent variables is obtained, thus determining the number of production lines already started in each discrete production process. The change in inventory of the product storage warehouse within a scheduling step. A set of solutions represents the vertices of these variables in the decision space; if not, the process continues until a solution is found. If the judgment matrix composed of the criteria is full rank, a feasible solution for a set of independent variables is obtained; therefore, this method searches all vertices in the variable space to determine the adjustable domain of the decision variables.

[0111] After completing the traversal, several feasible solutions with independent variables are obtained, which means the number of production lines that have been started in each discrete production process. Output of continuous production processes Output from flexible production processes The change in inventory of the product storage warehouse within a scheduling step. Given several sets of values ​​for the independent variables, the feasible solutions of each set of independent variables constitute the solution set of the multi-energy demand feasible solution. The feasible solutions of each set of independent variables are denoted as a matrix. subscript This indicates that the solution set is the solution set under a certain system state.

[0112] The feasible solutions for each set of independent variables in the feasible solution set of multi-energy demand are transformed into feasible solutions for various energy consumptions through a production process output and energy consumption relationship model, as follows:

[0113]

[0114] in, Indicates that the random variable is in the th position. The set of feasible solutions for multi-energy demand under discrete states; The transformation matrix representing the control variables of output and energy consumption in the production process can be obtained from the material and energy relationships in steps 1) and 2).

[0115] For all solutions in the feasible solution set of multi-energy demand, find the maximum and minimum values ​​of a single energy source to obtain the next... A multi-functional feasible interval for industrial load under discrete conditions.

[0116] 4) Based on the preset probability value requirement of the multi-energy flexibility of the industrial load, search for feasible intervals that meet the preset probability value requirement in the multi-energy feasible intervals under each discrete state, obtain the multi-energy flexibility of the industrial load, and thus realize the evaluation of the multi-energy flexibility of the industrial load.

[0117] In step 4), based on the preset probability value requirement for the multi-energy flexibility of the industrial load, a feasible interval that meets the preset probability value requirement is searched in the multi-energy feasible intervals under each discrete state to obtain the multi-energy flexibility of the industrial load, as follows:

[0118] In the entire industrial load There are n random variables, each random variable includes There are discrete states for the industrial load. One, according to step 3), obtain the industrial load. For each discrete state of the multi-energy feasible interval, the flexible adjustment interval of each energy is searched. For each energy, the maximum and minimum values ​​of the feasible interval of the energy in all discrete states are first obtained. Starting from the minimum value, the search calculation begins with a preset equal step size. The probability of each step size point is calculated, that is, the probability values ​​of the step size points in all discrete states are added together. Finally, after obtaining the probability values ​​of each step size point in the flexible adjustment interval of the energy, the flexible adjustment interval of the energy that meets the preset probability value requirements is selected, and finally the multi-energy flexibility of the industrial load is obtained.

[0119] Specific embodiments of the present invention are as follows:

[0120] Based on step 1), the entire production process of an air conditioning factory's industrial load in this embodiment can be divided into six production processes: a coil production line, a sheet metal production process, a two-piece forming process, a compressed air production process, an electronic component production process, and a final assembly process, along with their corresponding product storage warehouses: a sheet metal warehouse, a molded housing warehouse, a molded two-piece forming warehouse, an electronic component warehouse, and a compressed air storage tank. The types of these production processes and the parameters relating their energy consumption and output are shown in Table 1 below.

[0121] Table 1

[0122] Production process name Production process type Electricity consumption per unit output (kW) <![CDATA[Natural gas consumption per unit output (m 3 / h)]]> Heat consumption per unit output (kW) Compressed air consumption per unit output (kW) Product storage warehouse capacity Production requirements Compressed air production process Continuous production process 1 - - -1.1 <![CDATA[1240-1760 m 3 ]]> - Winding production line Discrete manufacturing process 0.93 - - - 2100 825 Sheet metal production process Continuous production process 0.71 0.057 - 0.57 1900 1176 Two-device processing production process Continuous production process 7.6 0.4 - 5 150 141 Electronic component manufacturing process Flexible production processes 7.14 - - - 30 21 Final assembly production process Flexible production processes 7.14 - 1.4 0.71 3600 3359

[0123] In addition, the parameters related to uncertainty in each production process are shown in Table 2 below:

[0124] Table 2

[0125] Production process name Minimum output and uncertainty Maximum output and uncertainty Remark Compressed air production process 0 <![CDATA[2280 m 3 / h]]> The compressed air unit is automatically managed and supplies compressed air to other production processes. It has high stability and does not consider uncertainties. Winding production line 0 3×70 This is a discrete manufacturing process with 3 production lines. Each production line has a unit output of 70. Start-up and shutdown are relatively deterministic, with 2 lines starting. Uncertainty is not considered. Sheet metal production process A half-normal distribution with a mean of 250 and a variance of 11. A half-normal distribution with a mean of 415 and a variance of 23. Two-device processing production process A half-normal distribution with a mean of 127 and a variance of 6. A half-normal distribution with a mean of 230 and a variance of 14. Electronic component manufacturing process A half-normal distribution with a mean of 88 and a variance of 4. A half-normal distribution with a mean of 165 and a variance of 9. Final assembly production process A half-normal distribution with a mean of 88 and a variance of 4. A half-normal distribution with a mean of 165 and a variance of 9.

[0126] Based on this, the relationship between energy consumption and output for each industrial process can be obtained from Table 1, and the random variables representing uncertainty in the output adjustment constraints can be obtained from Table 2.

[0127] Following step 2), graph theory is used to connect the various production processes and product storage warehouses. The connection results are as follows: Figure 2 As shown in Table 1, the parameters constraining the construction of the product storage warehouse are also listed. Additionally, they can be determined based on... Figure 2 The series and parallel relationships of each production process establish the output requirement constraints for each production process.

[0128] Following step 3), the parameters in Table 2 are discretized, with each random variable discretized into four states, resulting in 65,536 states for the entire industrial load. Solving for the multi-energy feasible interval in each state yields a set of feasible solutions. Each feasible solution in this set is then transformed into a feasible solution for multiple energy consumption scenarios using the production process's output and energy consumption relationships. By calculating the maximum and minimum values ​​of a single energy source across all solutions in this set, the multi-energy feasible interval for the industrial load under these system states can be obtained. Figure 3 This demonstrates the multi-energy feasible interval of industrial load under a certain system state, where all random variables take the mean value. In other words, if we disregard system uncertainties, the multi-energy flexible interval of industrial load is... Figure 3 As shown.

[0129] According to step 4), search for feasible intervals that meet the probability value requirements among all discrete state multi-energy feasible intervals to form the multi-energy flexibility of industrial load. Figure 4 It shows the industrial load multi-function adjustment range with a probability value of 90%.

[0130] As can be seen from the above embodiments, the final result obtained by the present invention evaluates the multi-energy regulation range of industrial load and the probability value of the range. Compared with the result that does not consider uncertainty ( Figure 3As shown in the results, the adjustable ranges for various energy sources have decreased, especially for electricity. Electricity regulation is currently the most needed flexibility capability in energy systems. Considering uncertainty, the high-probability regulation capability is reduced by about half compared to when uncertainty was not considered. This tells the energy system that when utilizing flexible resources, a portion of the industrial load's regulation capability is highly executable and can be readily utilized, while another portion may be unavailable in practice. The system can consider utilizing other regulation capabilities in non-emergency situations. Therefore, this invention clarifies the relationship between regulation capability and uncertainty, providing a reliable assessment of the energy system's regulation capability, and thus assisting the energy system in formulating various flexible operating schemes.

Claims

1. A method for assessing the multi-energy flexibility of industrial loads considering uncertainty, characterized in that: The method includes the following steps: 1) Construct a model of the energy consumption and output relationship of a production process under industrial load that considers output adjustment constraints; 2) Use graph theory to associate the production process and product storage warehouse in the industrial load of step 1), and construct the storage constraints of the product storage warehouse and the output requirement constraints of the production process. 3) Treat the uncertainties in the production process of industrial load as random variables, discretize the random variables to construct a set of discrete states of industrial load. Each set of discrete states includes several discrete states of industrial load. Based on the energy consumption and output relationship model, search the feasible interval for output adjustment constraints, storage constraints of product storage warehouses and output requirement constraints of production process under each discrete state to construct a multi-energy feasible interval under each discrete state. 4) Based on the preset probability value requirement of the multi-energy flexibility of the industrial load, search for feasible intervals that meet the preset probability value requirement in the multi-energy feasible intervals under each discrete state, obtain the multi-energy flexibility of the industrial load, and thus realize the evaluation of the multi-energy flexibility of the industrial load. In step 1), the production process in industrial load includes discrete production processes, continuous production processes, and flexible production processes. The energy consumption and output relationship model of the production process in industrial load is as follows: The energy consumption and output relationship model for discrete manufacturing processes is as follows: in, Indicates the output of a discrete manufacturing process; This indicates the number of production lines that have been activated during the discrete manufacturing process. This represents the output of a single production line within a discrete manufacturing process. This represents the total number of production lines in a discrete manufacturing process. This represents a column vector of various energy consumptions in a discrete production process, including the consumption of electrical energy, natural gas energy, and thermal energy. This represents a column vector of various energy consumptions for a single production line within a discrete production process. The column vector of various energy consumptions includes the consumption of electrical energy, natural gas energy, and thermal energy. The energy consumption and output relationship model for a continuous production process is as follows: in, This represents a column vector of various energy consumptions in a continuous production process, including the consumption of electrical energy, natural gas energy, and thermal energy. Indicates the output of a continuous production process; This indicates the relationship between various energy consumption and output in a continuous production process. The energy consumption and output relationship model for flexible production processes is as follows: in, This represents a column vector of various energy consumptions in a flexible production process, including the consumption of electrical energy, natural gas energy, and thermal energy. Indicates output from a flexible production process; Linear terms representing the relationship between energy consumption and output in flexible production processes; The constant term representing the relationship between energy consumption and output in flexible production processes; The specific production adjustment constraints are as follows: in, and These represent the minimum and maximum output of a discrete production process within a scheduling step, respectively. and These represent the minimum and maximum output of a flexible production process within a scheduling step, respectively. In step 3), based on the energy consumption and output relationship model, a feasible interval search is performed on the output adjustment constraint, the storage constraint of the product storage warehouse, and the output requirement constraint of the production process under each discrete state to construct a multi-energy feasible interval for each discrete state. Specifically, based on the output adjustment constraint in step 1) and the storage constraint of the product storage warehouse and the output requirement constraint of the production process in step 2), the vertex enumeration method is used to search for the feasible set of all random variables. After obtaining the feasible set of all random variables under each discrete state, the output and energy consumption relationship model of the production process in step 1) outputs the multi-energy feasible interval of the industrial load under each discrete state.

2. The method for assessing the multi-energy flexibility of industrial loads considering uncertainties according to claim 1, characterized in that: In step 2), graph theory is used to associate the production process and product storage warehouse in the industrial load of step 1). Specifically, the production process and product storage warehouse are modeled as nodes, and the material flow between the production process and product storage warehouse is the connection between the nodes.

3. The method for assessing the multi-energy flexibility of industrial loads considering uncertainties according to claim 1, characterized in that: In step 2), the storage constraints of the product storage warehouse are as follows: in, and These represent the maximum material outflow and material inflow within a scheduling step of the production process, respectively. and These represent the minimum and maximum material storage capacity of the product storage warehouse, respectively. This indicates the change in inventory of products stored in warehouses within one scheduling step under industrial load. This represents the existing inventory of the product storage warehouse within the industrial load at time t-1, prior to the scheduling cycle at time t.

4. The method for assessing the multi-energy flexibility of industrial loads considering uncertainties according to claim 2, characterized in that: In step 2), the specific output requirements of the production process are as follows: in, A column vector representing the change in inventory of all product storage warehouses within an industrial load over a scheduling step. This is a column vector representing the existing inventory of all product storage warehouses within the industrial load at time t-1 before the scheduling cycle at time t. This indicates the time of the entire scheduling cycle of the production process; A column vector representing the maximum output of all production processes; A relational matrix representing the production process and product storage warehouses; This represents a column vector of target output for each production process within a scheduling cycle.

5. The method for assessing the multi-energy flexibility of industrial loads considering uncertainties according to claim 4, characterized in that: The correlation matrix between the production process and the product storage warehouse. It includes several related elements, as follows: in, An association matrix representing the production process and product storage warehouses. The associated element in the i-th row and j-th column of the array; This represents the product conversion factor.

6. The method for assessing the multi-energy flexibility of industrial loads considering uncertainties according to claim 1, characterized in that: In step 3), under the constraints of storage in the product storage warehouse and output requirements in the production process, the random variables are discretized to construct a set of discrete states for the industrial load. For each set of discrete states... The details are as follows: in, and These represent the minimum and maximum output of the discrete production process within one scheduling step when the random variable is in its first discrete state; and These represent the minimum and maximum output of the flexible production process within a scheduling step when the random variable is in its first discrete state; and These represent the minimum and maximum output of the discrete production process within one scheduling step when the random variable is in its second discrete state; and These represent the minimum and maximum output of the flexible production process within one scheduling step when the random variable is in its second discrete state; and Each of the random variables in the industrial load represents the state at the _th ... For a discrete state, the minimum and maximum output of a discrete production process within a scheduling step; and Each of the random variables in the industrial load represents the state at the _th ... For a discrete state, the minimum and maximum output of a flexible production process within a scheduling step; and Each of the random variables in the industrial load represents the state at the _th ... For a discrete state, the minimum and maximum output of a discrete production process within a scheduling step; and Each of the random variables in the industrial load represents the state at the _th ... For a discrete state, the minimum and maximum output of a flexible production process within a scheduling step.

7. The method for assessing the multi-energy flexibility of industrial loads considering uncertainties according to claim 1, characterized in that: The vertex enumeration method specifically involves first denoting the number of discrete production processes in the industrial load as follows: The number of continuous production processes is denoted as The number of flexible production processes is denoted as The number of product storage warehouses is denoted as The independent variables in the production and energy consumption relationship model are: Then, the feasible interval is searched by traversing the production adjustment constraints in the industrial load, as follows: First fix The values ​​of the discrete production process variables are randomly selected from the output adjustment constraints in step 1), the storage constraints of the product storage warehouse in step 2), and the output requirement constraints of the production process. Constraints, determine by If the judgment matrix composed of the criteria is full rank, then a feasible solution for a set of independent variables is obtained, thus determining the number of production lines already started in each discrete production process. Changes in inventory of product storage warehouse within a scheduling step If not, continue iterating until the result is obtained. If the judgment matrix composed of the criteria is of full rank, a feasible solution with a set of independent variables is obtained; After completing the traversal, several feasible solutions with independent variables are obtained, which means the number of production lines that have been started in each discrete production process. Output of continuous production processes Output from flexible production processes Changes in inventory of product storage warehouse within a scheduling step Given several sets of values ​​for the independent variables, the feasible solutions of each set of independent variables constitute the solution set of the multi-energy demand feasible solution. The feasible solutions of each set of independent variables are denoted as a matrix. ; The feasible solutions for each set of independent variables in the feasible solution set of multi-energy demand are transformed into feasible solutions for various energy consumptions through a production process output and energy consumption relationship model, as follows: in, Indicates that the random variable is in the th position. The set of feasible solutions for multi-energy demand under discrete states; A transformation matrix representing the variables between control variables for output and variables for energy consumption in a production process; For all solutions in the feasible solution set of multi-energy demand, find the maximum and minimum values ​​of a single energy source to obtain the next... A multi-functional feasible interval for industrial load under discrete conditions.

8. The method for assessing the multi-energy flexibility of industrial loads considering uncertainties according to claim 1, characterized in that: In step 4), based on the preset probability value requirement for the multi-energy flexibility of the industrial load, a feasible interval that meets the preset probability value requirement is searched in the multi-energy feasible intervals under each discrete state to obtain the multi-energy flexibility of the industrial load, as detailed below: In the entire industrial load There are n random variables, each random variable includes There are discrete states for the industrial load. One, according to step 3), obtain the industrial load. For each discrete state of the multi-energy feasible interval, the flexible adjustment interval of each energy is searched. For each energy, the maximum and minimum values ​​of the feasible interval of the energy in all discrete states are first obtained. Starting from the minimum value, the search and calculation begin with a preset equal step size. The probability of each step size point is calculated, that is, the probability values ​​of the step size points in all discrete states are added together. Finally, after obtaining the probability values ​​of each step size point in the flexible adjustment interval of the energy, the flexible adjustment interval of the energy that meets the preset probability value requirements is selected, and the multi-energy flexibility of the industrial load is finally obtained.

Images