Power plant and electrolytic cell scheduling method considering multi-electrolytic cell production state and thermal coupling

Abstract

The application discloses a power plant and electrolytic cell scheduling method considering multiple electrolytic cell production states and thermal coupling, and relates to the technical field of source-load coordination control. The existing scheduling method has poor scheduling accuracy and influences the safety of electrolytic cells. An upper target model with the minimum total cost of a system as an objective is established, upper constraint conditions are established to constrain the upper target model, the upper target model is solved, the total power of an aluminum electrolysis cluster, the power of a thermal power plant and the production state of each electrolytic cell at each time in a future period are obtained, a lower target function with the minimum total cost as an objective is established, the target function is transformed by using a multiple electrolytic cell cluster power-temperature coupling model, a transformed lower target function is obtained, and lower constraint conditions are established to constrain the transformed lower target function. The power of each electrolytic cell at each time except an initial time is solved, and the power and the power of the thermal power plant are taken as a scheduling scheme. The application is used for realizing the scheduling of the thermal power plant and the electrolytic cell.

Description

Technical Field

[0001] This invention relates to the field of source-load coordination control technology. Background Technology

[0002] With the increasing proportion of new energy power generation such as wind power and photovoltaic power, the demand for peak-shaving resources in the power system is growing. As a traditional peak-shaving power source, thermal power units need to frequently adjust their output to balance the fluctuations of new energy sources. However, when thermal power units are deeply peak-shaving (output is lower than 50% or even 30% of rated power), they need to inject oil for combustion, which exacerbates equipment wear, increases emissions, and leads to a significant increase in operating costs.

[0003] Electrolytic aluminum, as a typical high-energy-consuming industrial load, features adjustable power, fast response, and large capacity, making it an ideal flexible regulation resource. Current research explores utilizing electrolytic aluminum loads for grid peak shaving. By adjusting the total power consumption of electrolytic aluminum plants, power can be increased when there is a surplus of new energy sources (increasing absorption and reducing downward pressure on thermal power), and decreased when there is a shortage of new energy sources (replacing upward pressure on thermal power). This aims to keep thermal power units within the primary peak-shaving economic range as much as possible, thereby reducing the overall peak-shaving cost of the system.

[0004] Currently, the scheduling method for using electrolytic aluminum to participate in grid peak shaving involves simplifying the electrolytic aluminum workshop into an equivalent single cell, establishing a thermal balance model, and predicting the total power of the electrolytic aluminum cluster and the status of thermal power units based on the output of new energy sources. However, simplifying the electrolytic aluminum workshop into an equivalent single cell ignores the thermal coupling between multiple electrolytic cells, the production status of each cell (calcination, electrode replacement, cell shutdown), and differences in specifications (rated power, capacity). This results in the obtained electrolytic aluminum scheduling scheme not matching the actual operating conditions, which can easily cause some electrolytic cells to exceed their temperature limits, affecting the safety of the electrolytic cells and leading to poor scheduling accuracy. Summary of the Invention

[0005] The purpose of this invention is to solve the problems of poor scheduling accuracy and impact on the safety of electrolytic cells in existing scheduling methods. It proposes a scheduling method for thermal power units and electrolytic cells that takes into account the production status and thermal coupling of multiple electrolytic cells.

[0006] A scheduling method for thermal power units and electrolytic cells that takes into account the production status and thermal coupling of multiple electrolytic cells, the method comprising:

[0007] Step 1: Collect the temperature of each electrolytic cell at the initial moment of the nth cycle, where the initial value of n is 1;

[0008] Step 2: Obtain the predicted values ​​of new energy power output and load power consumption at each time point within the nth cycle;

[0009] Step 3: Based on the predicted values ​​obtained in step 2, establish an upper-level objective model with the goal of minimizing the total system cost, and establish upper-level constraints to constrain the upper-level objective model. Use a mixed integer programming solver to solve the upper-level objective model to obtain the total power of the electrolytic aluminum cluster, the power of the thermal power unit, and the production status of each electrolytic cell at each time point in the nth cycle.

[0010] Step 4: Based on the total power of the electrolytic aluminum cluster and the production status of each electrolytic cell obtained in Step 3, establish a lower-level objective function with the goal of minimizing the total cost. Transform the objective function using the established multi-electrolytic cell cluster power-temperature coupling model to obtain the transformed lower-level objective function. Establish lower-level constraint conditions to constrain the transformed lower-level objective function.

[0011] Step 5: Input the temperature of each electrolytic cell at the initial moment of the nth cycle into the transformed lower objective function, and use a quadratic programming solver to solve the transformed lower objective function to obtain the power of each electrolytic cell at other times in the nth cycle except the initial moment. Use this power and the power of the thermal power unit obtained in Step 3 as the scheduling scheme for the nth cycle.

[0012] Step 6: Determine if n is equal to the preset scheduling cycle. If yes, stop. If no, let n = n + 1, input the temperature of each electrolyzer at each time in the (n-1)th cycle into the established MHE optimization model, solve for the temperature of each electrolyzer at the initial time of the nth cycle, and execute step 2.

[0013] The beneficial effects of this invention are:

[0014] This invention establishes a power-temperature coupled state-space model for a multi-electrolysis cell cluster. This model considers inter-cell thermal coupling, production state-dependent parameters of each cell, and differences in cell specifications. It also uses constraints to limit temperature and the number of adjustments to ensure production safety and equipment lifespan.

[0015] This invention designs an upper-level target model and a lower-level target model. The upper-level target model determines the total power of the cluster and the production status, while the lower-level target model allocates the power of each tank in real time and meets temperature safety constraints.

[0016] This invention deeply couples electrolytic aluminum load regulation with peak shaving of thermal power units. By optimizing the power variation of electrolytic aluminum load, it reduces the frequency and depth of thermal power units entering high-cost second-order peak shaving. Solving the upper and lower objective functions yields the output of thermal power units and the power of each electrolytic cell. This scheduling scheme enables the upper objective model to increase the total power of the electrolytic aluminum cluster (upward adjustment) when there is a surplus of renewable energy (high wind / solar power generation). Thermal power units can reduce their output to avoid entering second-order peak shaving. The lower objective model allocates the increased power to normal cells with large heat capacity and temperature margin, preventing small cells or roasting cells from exceeding limits. When renewable energy is insufficient (peak electricity demand or low wind season), the upper objective model optimizes and reduces (downward adjustment). Thermal power units reduce the upward output, saving fuel. The lower objective model allocates the reduced power to cells with limited production or normal cells, preventing the temperature of roasting cells and electrode switching cells from dropping too quickly. Therefore, this invention utilizes the model output of thermal power units and electrolytic cell power to achieve precise scheduling of electrolytic cells and thermal power units.

[0017] This invention uses the MHE optimization model to obtain the electrolytic cell temperature at the initial moment of each cycle except the first cycle. Compared with the measured initial moment temperature, the MHE optimization model has the function of correcting the temperature, making the initial moment temperature more accurate, and making the subsequent scheduling scheme more accurate. Attached Figure Description

[0018] Figure 1 A flowchart of a source-load coordination optimization method for thermal power peak shaving that takes into account the production status and thermal coupling of multiple electrolytic cells. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but this is not intended to limit the scope of the invention.

[0021] Example:

[0022] A source-load coordination optimization method for thermal power peak shaving, taking into account the production status and thermal coupling of multiple electrolytic cells, includes:

[0023] Step 1: Collect the temperature of each electrolytic cell at the initial moment of the nth cycle, where the initial value of n is 1;

[0024] Step 2: Obtain the predicted values ​​of new energy power output and load power consumption at each time point within the nth cycle;

[0025] Step 3: Based on the predicted values ​​obtained in step 2, establish an upper-level objective model with the goal of minimizing the total system cost, and establish upper-level constraints to constrain the upper-level objective model. Use a mixed integer programming solver to solve the upper-level objective model to obtain the total power of the electrolytic aluminum cluster, the power of the thermal power unit, and the production status of each electrolytic cell at each time point in the nth cycle.

[0026] Step 4: Based on the total power of the electrolytic aluminum cluster and the production status of each electrolytic cell obtained in Step 3, establish a lower-level objective function with the goal of minimizing the total cost. Transform the objective function using the established multi-electrolytic cell cluster power-temperature coupling model to obtain the transformed lower-level objective function. Establish lower-level constraint conditions to constrain the transformed lower-level objective function.

[0027] Step 5: Input the temperature of each electrolytic cell at the initial moment of the nth cycle into the transformed lower objective function, and use a quadratic programming solver to solve the transformed lower objective function to obtain the power of each electrolytic cell at other times in the nth cycle except the initial moment. Use this power and the power of the thermal power unit obtained in Step 3 as the scheduling scheme for the nth cycle.

[0028] Step 6: Determine if n is equal to the preset scheduling cycle. If yes, stop. If no, let n = n + 1, input the temperature of each electrolyzer at each time in the (n-1)th cycle into the established MHE optimization model, solve for the temperature of each electrolyzer at the initial time of the nth cycle, and execute step 2.

[0029] Specifically, the upper level makes decisions on the total power output of electrolytic aluminum to minimize the peak-shaving cost of thermal power, while the lower level allocates power in real time and ensures temperature safety, achieving a win-win situation for both the power grid and the load.

[0030] Further defined, the total system cost includes the power generation cost of thermal power units, the peak shaving cost of thermal power units, the cost of penalties for curtailment of renewable energy, the cost of load regulation and compensation for electrolytic aluminum, and the cost of switching production status.

[0031] The peak-shaving cost of thermal power units includes first-order peak-shaving cost and second-order peak-shaving cost. The second-order peak-shaving cost includes wear and tear cost, fuel consumption cost and additional emission pollution cost.

[0032] Specifically, peak shaving for thermal power units is divided into two stages:

[0033] First-order peak shaving: The unit output is within the normal peak shaving range, and the cost is mainly fuel cost.

[0034] Second-order peak shaving: When the unit output enters the deep peak shaving range, additional costs such as wear and tear, fuel consumption, and emissions pollution need to be considered.

[0035] Operating costs of thermal power units:

[0036] ,

[0037] In the formula, This represents the total operating cost of a thermal power unit; This indicates the cost of generating electricity from thermal power units; This indicates the peak-shaving cost of thermal power units; Indicates the time of thermal power unit Those who have made contributions.

[0038] The generation cost and peak-shaving cost are:

[0039]

[0040] ,

[0041] In the formula, Represents the coefficients of the quadratic function of power generation cost; These represent the first-order and second-order peak-shaving power thresholds, respectively. These represent wear and tear, fuel consumption, and additional emissions costs, respectively. This indicates a value that is higher than the cost of second-order peak shaving; the specific value needs to be defined based on the actual operating characteristics of the thermal power unit.

[0042] Further specifying, the upper-level target model is:

[0043] ,

[0044] In the formula, To minimize the total system cost, For the operating costs of thermal power units, Contribute to the practical development of new energy The resulting power generation costs, Forgone power The resulting costs of wind and solar power curtailment penalties The unit power adjustment compensation price for electrolytic aluminum load. This represents the change in the total power of the electrolytic aluminum cluster from the previous moment to the current moment. Costs associated with switching production status For indicator functions, for The production status of electrolytic aluminum at all times. This represents the number of time points included in the upper-level prediction time domain.

[0045] Specifically, in the day-ahead or intraday scheduling plan, the upper-level objective function aims to minimize the total system operating cost, coordinates the total power of thermal power units, new energy power generation, and electrolytic aluminum clusters, and simultaneously determines the production status plan of each electrolytic cell. It provides the lower-level real-time control with an economically optimal macro-instruction (i.e., total power baseline and production status).

[0046] In the formula, For electrolytic aluminum clusters at time Total active power absorbed from the power grid, This is the total power plan for the cluster provided by the upper layer.

[0047] Further restrictions are imposed on the upper-level constraints, including system power balance constraints, ramp rate constraints, spinning reserve constraints, peak-shaving phase transition logic constraints for thermal power units, state-output relationship constraints, minimum duration constraints, total power range constraints for electrolytic aluminum clusters, cluster temperature constraints, and temperature constraints.

[0048] Specifically, (1) System power balance:

[0049] ,

[0050] ,

[0051] In the formula: Indicates the wind turbine unit at time Those who have made meritorious contributions; Indicates the time of the photoelectric unit Those who have made meritorious contributions; Indicates at time The normal load power (excluding electrolytic aluminum).

[0052] (2) Upper and lower limits of thermal power unit output:

[0053] ,

[0054] In the formula: This indicates the upper and lower limits of the output of thermal power units.

[0055] (3) Slope rate constraint:

[0056] ,

[0057] In the formula: Indicates the uphill / downhill ramp rate limit.

[0058] (3) Rotational spare constraint:

[0059] The system needs to reserve sufficient upswing and downswing capacity to cope with fluctuations in new energy sources or sudden load changes. For a single thermal power unit, its available reserve capacity is limited by its current output and ramp rate.

[0060] ,

[0061] ,

[0062] In the formula: This indicates the required spin-up / rotation-down reserve capacity of the system; Indicates the duration of a time period.

[0063] (4) Peak-shaving phase switching logic for thermal power units (to avoid frequent switching):

[0064] Introducing binary variables and The definition is as follows:

[0065] Indicates the time period of the unit In first-order peak-shaving state Otherwise, it is 0.

[0066] This indicates that the unit is in second-order peak-shaving mode. Otherwise, it is 0.

[0067] in, These are the power thresholds for first-order and second-order peak modulation, respectively.

[0068] (4) Constraints on the relationship between state and output:

[0069] ,

[0070] ,

[0071] ,

[0072] ,

[0073] In the formula, Represents a large positive number; It represents a small positive number.

[0074] The above constraints guarantee that: when hour, ;when hour, Similarly, for second-order peak modulation.

[0075] Furthermore, second-order peak regulation always occurs within the first-order peak regulation range; therefore:

[0076] ,

[0077] (5) Minimum duration constraint:

[0078] To avoid frequent entry and exit of generating units from deep peak shaving ranges, it is stipulated that once a second-order peak shaving state is entered, it must maintain at least continuous operation. It can only exit after a certain time period. This can be achieved using the following linear inequality:

[0079] ,

[0080] when (That is, when it just enters the second-order peak modulation), the right end is Forced Future Within a time period All are 1.

[0081] When there is no change in state, the right side is 0, and the inequality is automatically satisfied.

[0082] In the formula: This represents the minimum duration of the second-order peak-shaving state of a thermal power unit.

[0083] (6) Total power range of electrolytic aluminum clusters:

[0084] ,

[0085] in , .

[0086] In the formula, Indicates the upper and lower limits of the power of the electrolytic aluminum cluster; These represent electrolytic cells. In a given production state The minimum and maximum active power allowed.

[0087] (7) Cluster temperature constraints

[0088] Equivalent single-slot model:

[0089] ,

[0090] in , After discretization:

[0091] ,

[0092] In the formula: This represents the equivalent heat capacity of the equivalent cluster. Cluster equivalent temperature represents the average thermal state of the entire electrolytic aluminum workshop; This represents the total heat consumption of the cluster's production. Indicates the cluster's equivalent heat dissipation coefficient; Indicates ambient temperature.

[0093] (8) Temperature constraint:

[0094] ,

[0095] In the formula: This indicates the upper and lower limits of the equivalent cluster temperature.

[0096] By rapidly adjusting the load of electrolytic aluminum, the output of thermal power units can be kept within the first-order peak-shaving range (P_thermal ≥ P_peak1) as much as possible, reducing the duration and frequency of second-order peak shaving. This increases the number of first-order peak shavings, reduces the number of second-order peak shavings, and lowers the overall peak-shaving cost.

[0097] Further specifying, the lower-level objective function is:

[0098] ,

[0099] In the formula, This indicates minimizing the total cost. For the number of time periods, Represents the predicted temperature vector. Represents the reference temperature vector. This represents the power increment vector for each electrolytic cell. Temperature uniformity coefficient, Electrolytic cell temperature, Electrolytic cell temperature, This indicates the tracking upper-level baseline coefficient. This indicates the total power plan for the cluster given by the upper layer. This represents the temperature tracking weight matrix. This is the power adjustment penalty weight matrix.

[0100] Further defining the power-temperature coupling model for a multi-electrolytic cell cluster is as follows:

[0101] ,

[0102] In the formula, express The column vector formed by the temperature increments of all electrolytic cells at any given time. Represents the state transition matrix. express The column vector formed by the temperature increments of all electrolytic cells at any given time. , for Temperature at any moment for The steady-state operating point temperature at all times. Represents the input matrix, Represents the power increment matrix, , for The power of the electrolytic cell at all times. for The steady-state operating point power of the time-point decomposition tank.

[0103] Specifically, the process of establishing the power-temperature coupling model for a multi-electrolytic cell cluster is as follows:

[0104] 1. Heat balance equation for a single electrolytic cell (basic):

[0105] For any electrolytic cell Its heat balance equation is:

[0106]

[0107] In the formula: Indicates the total heat capacity of the electrolyte; express Constant electrolytic cell temperature; Indicates the current time of the electrolytic cell; express The input power of the electrolytic cell at any given time; express The heat power consumed in the electrochemical production of electrolytic aluminum at all times; express The heat dissipation power of electrolytic aluminum at all times.

[0108] Heat dissipation employs Newton's law of cooling and linearizes the radiation term:

[0109]

[0110] In the formula: Indicates the environmental heat dissipation coefficient; Indicates ambient temperature.

[0111] 2. Introduction of inter-slot thermal coupling term:

[0112] There is heat exchange (radiation, conduction, flue gas flow) between adjacent electrolytic cells, assumed to be proportional to the temperature difference:

[0113]

[0114] In the formula: Indicates electrolytic cell With electrolytic cell The heat exchange power between them, when When, it indicates that heat is transferred from the tank. Flow channel Conversely, from the groove Flow channel ; Indicates slot With slot The thermal coupling coefficient between the tanks comprehensively reflects factors such as the heat exchange area between the tanks, the distance, the thermal resistance of the insulation material, and the radiation angle coefficient. The larger the coefficient, the more heat is exchanged under the same temperature difference. , This indicates the temperature of different electrolytic cells.

[0115] The complete heat balance equation is:

[0116]

[0117] In the formula: express Electrolytic cell heat capacity; express Electrolytic cell Temperature at any given time; express Electrolytic cell The input power of the electrolytic cell at any given time; express Electrolytic cell The heat power consumed in the electrochemical production of electrolytic aluminum at all times; express Electrolytic cell environment heat dissipation coefficient.

[0118] 3. The impact of production status on model parameters:

[0119] Production status of each electrolytic cell , Under different states, the model parameters are taken as the baseline value multiplied by the state correction factor:

[0120]

[0121]

[0122]

[0123]

[0124] In the formula: express Electrolytic cells in current production status The actual heat capacity below; express The baseline heat capacity of an electrolytic cell under normal production conditions is determined by its specifications (rated power, capacity, etc.); Production status correction factor for heat capacity; express Electrolytic cells in current production status The equivalent environmental heat dissipation coefficient is as follows; express The baseline heat dissipation coefficient of an electrolytic cell under normal production conditions is determined by its heat dissipation area, surface emissivity, and other specifications. The production state correction factor represents the heat dissipation coefficient. express Electrolytic cells in current production status The heat consumption power of electrochemical production under these conditions; express The baseline production heat consumption of an electrolytic cell under normal operating conditions is usually related to its rated power and energy efficiency. A production state correction factor indicating the heat consumption of production; Indicates electrolytic cell and In the current production status , Thermal coupling coefficient under; The reference thermal coupling coefficient represents the condition under normal production conditions for both tanks, and is determined by specifications such as adjacent side area, spacing, and insulation conditions. This represents the production state joint correction factor for the thermal coupling coefficient.

[0125] Production state parameterization: Industrial states such as calcination, electrode replacement, and tank shutdown are quantified into correction factors such as heat capacity, heat dissipation coefficient, and power adjustment range, so that the model can adapt to different production stages.

[0126] 4. Modeling of differences in electrolytic cell specifications:

[0127] Different electrolytic cells have different inherent specifications such as rated power, capacity, and heat dissipation area, which affect the baseline parameters.

[0128] (1) Reference heat capacity:

[0129]

[0130] In the formula: This indicates the average specific heat capacity of the material (molten aluminum, electrolyte, etc.). express The total mass of materials participating in heat exchange inside the electrolytic cell; Indicates the proportionality coefficient fitted from experimental or design data; express The rated active power of the electrolytic cell.

[0131] Specification difference modeling: Based on the inherent specifications such as rated power and capacity of each tank, establish the proportional relationship of the benchmark parameters to make the model applicable to mixed tank workshops.

[0132] (2) Reference heat dissipation coefficient:

[0133]

[0134] In the formula: The convective heat transfer coefficient depends on the airflow conditions at the surface of the tank and can be considered a constant. express The total heat dissipation area of ​​the electrolytic cell includes the side walls, top surface, and bottom surface; express The total heat dissipation area of ​​the electrolytic cell includes the side walls, top surface, and bottom surface; It is the Stefan-Boltzmann constant; The emissivity of the tank shell surface is related to the material and the degree of oxidation. This indicates the effective radiation area, which is usually slightly smaller than the total heat dissipation area. express The rated operating temperature of the electrolytic cell; This represents the ratio of the heat dissipation coefficient to the heat dissipation area. It represents the ratio of the heat dissipation coefficient to the power of 2 / 3 of the rated power.

[0135] (3) Coupling coefficient between reference slots:

[0136]

[0137] In the formula: Indicates electrolytic cell and The reference thermal coupling coefficient (W / K) under normal production conditions represents the heat exchange power per unit temperature difference; The reference scaling factor (W / (m·K)) representing the thermal coupling coefficient is determined by a combination of factors such as the heat insulation conditions of the slot and the radiation angle coefficient. This represents the minimum area of ​​the adjacent sides of the two tanks. Taking the smaller value means that the actual effective heat exchange area is limited to the smaller side. Indicates electrolytic cell and The center distance or the distance between adjacent faces; Simplified proportionality coefficient; express The rated active power of the electrolytic cell.

[0138] (4) Power adjustment range:

[0139]

[0140]

[0141] In the formula: Indicates electrolytic cell In the current production status The minimum permissible active power; Indicates electrolytic cell In the current production status The maximum permissible active power; Indicates production status The corresponding minimum power ratio coefficient; Indicates production status The corresponding maximum power ratio coefficient.

[0142] (5) Incremental form and state-space model:

[0143] At steady-state operating point Near-linearization, satisfying the following in steady state:

[0144]

[0145] In the formula: Indicates electrolytic cell Stable power; Indicates electrolytic cell steady-state temperature; Indicates electrolytic cell The steady-state temperature.

[0146] Define increment , Substituting, we get:

[0147]

[0148] In the formula: Indicates electrolytic cell Temperature increment; Indicates electrolytic cell The power increment; Indicates electrolytic cell Temperature increment.

[0149] Written in matrix form:

[0150]

[0151] in:

[0152] ,

[0153]

[0154] ,

[0155] In the formula: This represents a column vector consisting of the temperature increments of all electrolytic cells. This represents a column vector consisting of the power increments of all electrolytic cells; This represents a diagonal matrix of heat capacity, where the diagonal elements are the actual heat capacity of each slot. Represents the coupling matrix. Represents the diagonal elements in a matrix. This represents the off-diagonal elements in the matrix.

[0156] Discretization (forward Euler, sampling time) ):

[0157]

[0158] In the formula: This represents a column vector consisting of the temperature increments of all electrolytic cells; This represents a column vector consisting of the temperature increments of all electrolytic cells at the next moment; Represents the identity matrix; express The inverse of the heat capacity diagonal matrix; An aggregation matrix representing the production status; This represents the power increment matrix.

[0159] definition:

[0160] ,

[0161] In the formula: Represents the state transition matrix, describing the transition from... arrive A linear mapping; Represents the input matrix, describing right The impact.

[0162] The final discrete state-space model is:

[0163]

[0164] The model is a time-varying linear system, and its parameters depend on the current production status and specification baseline of each tank.

[0165] Further restrictions are imposed on the lower-level constraints, including temperature safety constraints, power allocation and range constraints, total power balance constraints of the cluster, cumulative constraints on the number of adjustments, and power change rate constraints.

[0166] Specifically, the state update equation is: (multi-slot coupled model, incremental form)

[0167] ,

[0168] ,

[0169] ,

[0170] ,

[0171] In the formula: Indicates time The temperature increment vector has dimensions of ; This represents the time-varying state transition matrix, with dimension 1. This describes the transmission mechanism of temperature increments themselves. Indicates time The production state vector of each electrolytic cell at that time; Represents the time-varying input matrix with dimension . This maps the power increment to the temperature increment. Indicates time The power increment vector; This represents the steady-state temperature vector, and the rated operating temperature of each tank. Indicates the first The absolute temperature vector at each moment; Indicates the first The power vector at each time step; This represents the steady-state power vector, typically taken as the rated power. Represents the identity matrix; This represents the coupling matrix.

[0172] Initial state:

[0173] ,

[0174] In the formula: express The average temperature of the electrolytic aluminum cluster at any given time.

[0175] (2) Temperature safety constraints

[0176] ,

[0177] In the formula: These represent electrolytic cells. The minimum and maximum permissible temperatures are determined by process safety. Indicates electrolytic cell The steady-state temperature, i.e., the rated operating point temperature; Indicates electrolytic cell At discrete time Temperature increment.

[0178] (3) Power allocation and range constraints

[0179] ,

[0180] in: , , , These represent electrolytic cells. During the period The minimum active power allowed at that time. Indicates electrolytic cell At any moment The actual active power; Indicates the electrolytic aluminum cluster at time The total active power absorbed from the grid is determined by the upper-level optimization. Indicates production status The corresponding minimum power ratio coefficient; Indicates production status The corresponding maximum power ratio coefficient.

[0181] (4) Cluster total power balance (implicit in the total power definition, no additional constraints required)

[0182] ,

[0183] (5) Cumulative constraint on the number of adjustments (sliding window)

[0184] Define binary variables Indicates electrolytic cell At any moment Did an effective adjustment occur?

[0185] ,

[0186] definition ,when Conversely, it is 0. We can obtain:

[0187] ,

[0188] In the formula: Indicates electrolytic cell At any moment Active power; This represents the threshold of power change that can be effectively regulated. This represents a Boolean variable (with a value of 0 or 1), and indicates an electrolytic cell. At any moment Did an effective adjustment occur (typically when the power change exceeds a threshold)? (Time is recorded as 1). Indicates electrolytic cell At any moment The change in power relative to the previous moment; Electrolytic cell In production status The maximum number of adjustments allowed within a single scheduling cycle.

[0189] Sliding window cumulative constraint:

[0190] ,

[0191] In the formula: Indicates the width of the sliding window; Indicates the summation index; This indicates the maximum number of adjustments allowed within each sliding window. The value varies depending on the production status.

[0192] (6) Power change rate (slope constraint)

[0193] ,

[0194] In the formula: These represent electrolytic cells. The maximum allowable downward and upward power at any given moment.

[0195] Further specifying, the production state includes roasting state, electrode switching state, and cell shutdown state.

[0196] Further defining the total cost, it includes penalties for temperature tracking deviation, power regulation amplitude, inter-slot temperature difference balancing, and baseline deviation for tracking the upper-level total power.

[0197] Specifically, temperature balance control: a penalty term for temperature difference between tanks is introduced into the MPC objective function to actively suppress local overheating / overcooling and ensure consistent product quality.

[0198] Further specifying, the MHE optimization model is expressed as:

[0199] ,

[0200] In the formula, This indicates the estimated electrolytic cell temperature; Represents the process noise vector; This represents its weight matrix; Represents the measurement noise vector; It is represented as its weight matrix; This indicates the temperature of each electrolytic cell at any given moment within the historical period; This represents the initial state covariance weights.

[0201] Specifically, the cell temperature at the initial moment of the first cycle was measured, while the cell temperature at the initial moment of each subsequent cycle was obtained using the MHE optimization model. The specific calculation method was based on past measurements. The measured temperature of the electrolyzer at each moment is used to infer the current temperature of the electrolyzer. The constraints of the MHE optimization model are as follows:

[0202]

[0203] In the formula: This represents the prior estimate (usually the predicted value from the previous round of MPC). Indicates the initial state covariance weights; This represents the output matrix, typically an identity matrix (temperature is directly measurable). Solving for this matrix yields the current temperature of the electrolytic cell. As a power-temperature coupling model of multi-electrolyte clusters .

[0204] While the invention has been described herein with reference to specific embodiments, it should be understood that these embodiments are merely examples of the principles and applications of the invention. Therefore, it should be understood that many modifications can be made to the exemplary embodiments, and other arrangements can be designed without departing from the spirit and scope of the invention as defined by the appended claims. It should be understood that different dependent claims and features described herein can be combined in ways different from those described in the original claims. It is also understood that features described in conjunction with individual embodiments can be used in other described embodiments.

Claims

1. A scheduling method for thermal power units and electrolytic cells considering the production status and thermal coupling of multiple electrolytic cells, characterized in that, The method includes: Step 1: Collect the temperature of each electrolytic cell at the initial moment of the nth cycle, where the initial value of n is 1; Step 2: Obtain the predicted values ​​of new energy power output and load power consumption at each time point within the nth cycle; Step 3: Based on the predicted values ​​obtained in step 2, establish an upper-level objective model with the goal of minimizing the total system cost, and establish upper-level constraints to constrain the upper-level objective model. Use a mixed integer programming solver to solve the upper-level objective model to obtain the total power of the electrolytic aluminum cluster, the power of the thermal power unit, and the production status of each electrolytic cell at each time point in the nth cycle. Step 4: Based on the total power of the electrolytic aluminum cluster and the production status of each electrolytic cell obtained in Step 3, establish a lower-level objective function with the goal of minimizing the total cost. Transform the objective function using the established multi-electrolytic cell cluster power-temperature coupling model to obtain the transformed lower-level objective function. Establish lower-level constraint conditions to constrain the transformed lower-level objective function. Step 5: Input the temperature of each electrolytic cell at the initial moment of the nth cycle into the transformed lower objective function, and use a quadratic programming solver to solve the transformed lower objective function to obtain the power of each electrolytic cell at other times in the nth cycle except the initial moment. Use this power and the power of the thermal power unit obtained in Step 3 as the scheduling scheme for the nth cycle. Step 6: Determine if n is equal to the preset scheduling cycle. If yes, stop. If no, let n = n + 1, input the temperature of each electrolyzer at each time in the (n-1)th cycle into the established MHE optimization model, solve for the temperature of each electrolyzer at the initial time of the nth cycle, and execute step 2.

2. The scheduling method for thermal power units and electrolytic cells considering the production status and thermal coupling of multiple electrolytic cells according to claim 1, characterized in that, The total system cost includes the power generation cost of thermal power units, the peak shaving cost of thermal power units, the cost of penalties for curtailment of renewable energy, the cost of load regulation compensation for electrolytic aluminum, and the cost of switching production status. The peak-shaving cost of thermal power units includes first-order peak-shaving cost and second-order peak-shaving cost. The second-order peak-shaving cost includes wear and tear cost, fuel consumption cost and additional emission pollution cost.

3. The scheduling method for thermal power units and electrolytic cells considering the production status and thermal coupling of multiple electrolytic cells according to claim 1, characterized in that, The upper-level target model is: ， In the formula, To minimize the total system cost, For the operating costs of thermal power units, Contribute to the practical development of new energy The resulting power generation costs, Forgone power The resulting costs of wind and solar power curtailment penalties The unit power adjustment compensation price for electrolytic aluminum load. This represents the change in the total power of the electrolytic aluminum cluster from the previous moment to the current moment. Costs associated with switching production status For indicator functions, for The production status of electrolytic aluminum at all times. This represents the number of time periods included in the upper-level prediction time domain.

4. The scheduling method for thermal power units and electrolytic cells considering the production status and thermal coupling of multiple electrolytic cells according to claim 3, characterized in that, The upper-level constraints include system power balance constraints, ramp rate constraints, spinning reserve constraints, peak-shaving phase transition logic constraints for thermal power units, state-output relationship constraints, minimum duration constraints, total power range constraints for electrolytic aluminum clusters, cluster temperature constraints, and temperature constraints.

5. The scheduling method for thermal power units and electrolytic cells considering the production status and thermal coupling of multiple electrolytic cells according to claim 1 or 4, characterized in that, The lower-level objective function is: ， In the formula, This indicates minimizing the total cost. For the number of time periods, Represents the predicted temperature vector. Represents the reference temperature vector. This represents the power increment vector for each electrolytic cell. Temperature uniformity coefficient, Electrolytic cell temperature, Electrolytic cell temperature, This indicates the tracking upper-level baseline coefficient. This indicates the total power plan for the cluster given by the upper layer. This represents the temperature tracking weight matrix. This is the power adjustment penalty weight matrix.

6. The scheduling method for thermal power units and electrolytic cells considering the production status and thermal coupling of multiple electrolytic cells according to claim 5, characterized in that, The power-temperature coupling model for a multi-electrolysis cell cluster is as follows: ， In the formula, express The column vector formed by the temperature increments of all electrolytic cells at any given time. Represents the state transition matrix. express The column vector formed by the temperature increments of all electrolytic cells at any given time. , for Temperature at any moment for The steady-state operating point temperature at all times. Represents the input matrix, Represents the power increment matrix, , for The power of the electrolytic cell at all times. for The steady-state operating point power of the time-point decomposition tank.

7. The scheduling method for thermal power units and electrolytic cells considering the production status and thermal coupling of multiple electrolytic cells according to claim 6, characterized in that, The lower-level constraints include temperature safety constraints, power allocation and range constraints, total power balance constraints of the cluster, cumulative adjustment times constraints, and power change rate constraints.

8. The scheduling method for thermal power units and electrolytic cells considering the production status and thermal coupling of multiple electrolytic cells according to claim 1, characterized in that, The production states include roasting state, electrode switching state, and cell shutdown state.

9. The scheduling method for thermal power units and electrolytic cells considering the production status and thermal coupling of multiple electrolytic cells according to claim 1, characterized in that, The total cost includes penalties for temperature tracking deviation, power regulation amplitude, inter-slot temperature difference equalization, and baseline deviation for tracking the upper total power.

10. The scheduling method for thermal power units and electrolytic cells considering the production status and thermal coupling of multiple electrolytic cells according to claim 1, 7, 8 or 9, characterized in that, The MHE optimization model is expressed as: ， In the formula, This indicates the estimated electrolytic cell temperature; Represents the process noise vector; This represents its weight matrix; Represents the measurement noise vector; It is represented as its weight matrix; This indicates the temperature of each electrolytic cell at any given moment within the historical period; This represents the initial state covariance weights.

Images