Power control scheduling method and device based on prediction error, terminal equipment and storage medium

Abstract

The invention discloses an electric power control scheduling method and device based on a prediction error, terminal equipment and a storage medium. According to actual first power state values and second power state values of a new energy generator set and a thermal power generating set in an electric power system in a current time period, the prediction error is calculated; constructing an initial state space for representing the stable state of the power system according to the initial control strategy parameters and the predicted error state values of the new energy generator sets so as to accurately evaluate the stable state of the power system, and constructing an initial action space according to the initial action of each generator set regulated and controlled under the initial control strategy parameters; and iteratively optimizing the initial control strategy parameters according to the initial state space and the initial action space to generate target control strategy parameters capable of eliminating system fluctuation caused by prediction errors and enabling the power system to recover stability, and scheduling a new energy generator set and a thermal power generating unit. And the stable operation of the power system is effectively ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of power dispatching, and in particular, to a power control dispatching method, device, terminal device and storage medium based on prediction error. Background Art

[0002] The large-scale development and grid connection of wind power have caused significant changes in the power source structure of the traditional power system dominated by coal power. The randomness, volatility and unpredictability of wind power pose severe challenges to the safe operation of the power system. In this context, improving the power prediction accuracy of new energy generating units such as wind power and photovoltaic power is a positive measure, so that the power control department can reasonably arrange control plans based on the power prediction information of wind power and photovoltaic power, and then ensure the stable operation of the power system.

[0003] However, since prediction error is inevitable, it leads to the irrationality of the formulated control dispatching plan, causes fluctuations in the power system operating based on the dispatching plan, and then increases the additional cost of coping with the fluctuations. With the increase in the number of grid-connected wind farms, the uncertainty of wind power increases significantly. How to eliminate the power fluctuations caused by prediction error has become an urgent problem to be solved. Summary of the Invention

[0004] Embodiments of the present invention provide a power control dispatching method, device, terminal device and storage medium based on prediction error, which can eliminate the power fluctuations caused by prediction error and effectively ensure the stable operation of the power system.

[0005] An embodiment of the present invention provides a power control dispatching method based on prediction error, including:

[0006] Obtaining a first power state value of a new energy generating unit in the power system at the current time period, a second power state value of a thermal power generating unit at the current time period, and a predicted power state value of the new energy generating unit predicted at the current time period;

[0007] Calculating a prediction error state value according to the first power state value and the predicted power state value;

[0008] Constructing an initial state space for characterizing the operating state of the power system according to the first power state value, the second power state value, and the prediction error state value, and constructing an initial action space according to the change trends of the first power state value, the second power state value, and the prediction error state value under the initial control strategy parameters; wherein, the initial control strategy parameters are the control strategy parameters generated by the power system affected by prediction error at the current time period;

[0009] Iteratively optimize the initial control strategy parameters according to the initial state space and the initial action space, and evaluate the stable state of the power system according to the updated state space during the iteration. When it is determined that the power system operates stably under the target control strategy parameters generated during the current iteration according to the stable state, output the target control strategy parameters;

[0010] Perform scheduling control on the new energy generating units and the thermal power generating units according to the target control strategy parameters.

[0011] Further, constructing an initial state space for characterizing the operating state of the power system according to the first power state value, the second power state value, and the prediction error state value includes:

[0012] Obtain the load state values of several load nodes, the power flow state values of several branches, and the charge and discharge power state values of energy storage devices in the power system;

[0013] Calculate the power balance state value of the power system according to the first power state value, the second power state value, the charge and discharge power state value, and the load state value;

[0014] Detect the out-of-limit conditions of the power flow state values of each branch according to the upper limit and lower limit of the power flow of several branches, and generate the upper limit detection state value and the lower limit detection state value of the power flow of each branch;

[0015] Construct the initial state space according to the first power state value, the second power state value, the power flow state value, the power balance state value, the upper limit detection state value of the power flow, and the lower limit detection state value of the power flow.

[0016] Further, constructing an initial action space according to the change trends of the first power state value, the second power state value, and the prediction error state value under the initial control strategy parameters includes:

[0017] Construct a power action vector according to the change trends of the first power state value, the second power state value, and the prediction error state value under the initial control strategy parameters;

[0018] When it is determined that the power system is power balanced according to the power balance state value, and it is determined that no power flow out-of-limit occurs in each branch of the power system according to the upper limit detection state value of the power flow and the lower limit detection state value of the power flow, construct the initial action space according to the power action vector and the prediction error action vector.

[0019] Further, after constructing the prediction error action vector, it further includes:

[0020] When it is determined that there is a power imbalance in the power system and there is no power flow over-limit in each branch of the power system, a power balance action vector is constructed according to the change trend of the power balance state value under the initial control strategy parameters, and an initial action space is constructed according to the power balance action vector, the power action vector, and the prediction error action vector;

[0021] When it is determined that the power system is power balanced and there is a power flow over-limit in each branch of the power system, an upper limit action vector and a lower limit action vector are constructed according to the change trends of the upper limit detection state value and the lower limit detection state value under the initial control strategy parameters, and an initial action space is constructed according to the upper limit action vector, the lower limit action vector, the power action vector, and the prediction error action vector;

[0022] When it is determined that there is a power imbalance in the power system and there is a power flow over-limit in each branch of the power system, a power balance action vector is constructed according to the change trend of the power balance state value under the initial control strategy parameters, an upper limit action vector and a lower limit action vector are constructed according to the change trends of the upper limit detection state value and the lower limit detection state value under the initial control strategy parameters, and an initial action space is constructed according to the power balance action vector, the upper limit action vector, the lower limit action vector, the power action vector, and the prediction error action vector.

[0023] Further, the constructing of the power balance action vector according to the change trend of the power balance state value under the initial control strategy parameters includes:

[0024] When the power generation amount of the power system characterized by the power balance state value is greater than the load amount, it is determined that the first change trend constraint of the first power state value, the second power state value, and the discharge power state value of the energy storage device is negative, and the second change trend constraint of the charging power state value of the energy storage device is positive;

[0025] When the power generation amount of the power system characterized by the power balance state value is less than the load amount, it is determined that the first change trend constraint is positive and the second change trend constraint is negative;

[0026] The first change trend constraint and the second change trend constraint are used to construct a power balance action constraint;

[0027] According to the power balance action constraint and the change trend of the power balance state value under the initial control strategy parameters, a power balance action vector is constructed.

[0028] Preferably, constructing an upper limit action vector and a lower limit action vector according to the change trends of the upper limit detection status value and the lower limit detection status value of the power flow under the initial control strategy parameters includes:

[0029] According to the upper limit detection status value of the power flow and the lower limit detection status value of the power flow, determine the first branch where the power flow state value exceeds the upper limit of the power flow, and the second branch where the power flow state value exceeds the lower limit of the power flow;

[0030] Set the third change trend constraint of the power flow power flowing to the first branch when the new energy generating set, the thermal power generating set, and the energy storage device discharge to a negative value, and set the fourth change trend constraint of the charging power absorbed by the energy storage device from the first branch to a positive value;

[0031] Set the third change trend constraint of the power flow power flowing to the second branch when the new energy generating set, the thermal power generating set, and the energy storage device discharge to a positive value, and set the fourth change trend constraint of the charging power absorbed by the energy storage device from the second branch to a negative value;

[0032] Construct a power flow transfer constraint according to the third change trend constraint and the fourth change trend constraint of each branch;

[0033] Construct an upper limit action vector and a lower limit action vector according to the power flow transfer constraint, the change trend of the upper limit detection status value of the power flow, and the change trend of the lower limit detection status value of the power flow under the initial control strategy parameters.

[0034] Further, iteratively optimizing the initial control strategy parameters according to the initial state space and the initial action space, evaluating the stable state of the power system according to the updated state space during the iteration process, and when determining that the power system operates stably under the target control strategy parameters generated during the current iteration process according to the stable state, outputting the target control strategy parameters includes:

[0035] Repeatedly perform parameter optimization operations according to the initial state space and the initial action space until target control strategy parameters are generated;

[0036] Among them, the parameter optimization operation includes:

[0037] Obtain the strategy parameters to be optimized, the state space to be evaluated, and the action space to be evaluated. Among them, initially, the strategy parameters to be optimized are the initial control strategy parameters, the state space to be evaluated is the initial state space, and the action space to be evaluated is the initial action space;

[0038] Update the to-be-evaluated state space according to the to-be-evaluated action space, and generate an updated state space for characterizing the operating state of the power system after power dispatching according to the initial control strategy parameters;

[0039] Calculate the first steady-state score of the power system according to the to-be-evaluated state space, and calculate the second steady-state score of the power system according to the updated state space;

[0040] If the second steady-state score does not meet the preset score threshold, then use the advantage function to calculate the reward value according to the first steady-state score and the second steady-state score;

[0041] Update the optimization trend of the to-be-optimized strategy parameters according to the reward value, and update the to-be-optimized strategy parameters according to the optimization trend to generate updated strategy parameters;

[0042] Generate the to-be-evaluated action space required for the next round of parameter optimization operation according to the updated state space and the updated strategy parameters;

[0043] Use the updated strategy parameters and the updated state space as the to-be-optimized strategy parameters and the to-be-evaluated state space required for the next round of parameter optimization operation respectively, and perform the next round of parameter optimization operation;

[0044] If the second steady-state score meets the preset score threshold, then use the to-be-optimized strategy parameters as the target control strategy parameters.

[0045] Another embodiment of the present invention provides a power control and dispatching device based on prediction error, including:

[0046] A data acquisition module for acquiring the first power state value of the new energy generating unit in the power system at the current time period, the second power state value of the thermal power unit at the current time period, and the predicted power state value of the new energy generating unit predicted at the current time period;

[0047] An error calculation module for calculating the prediction error state value according to the first power state value and the predicted power state value;

[0048] A space construction module for constructing an initial state space for characterizing the operating state of the power system according to the first power state value, the second power state value, and the prediction error state value, and constructing an initial action space according to the change trend of the first power state value, the second power state value, and the prediction error state value under the initial control strategy parameters; wherein, the initial control strategy parameters are the control strategy parameters generated by the power system affected by the prediction error at the current time period;

[0049] A parameter optimization module, configured to iteratively optimize the initial control strategy parameters according to the initial state space and the initial action space, evaluate the stable state of the power system according to the updated state space during the iteration process, and output the target control strategy parameters when it is determined that the power system operates stably under the target control strategy parameters generated during the current iteration process according to the stable state;

[0050] A power dispatch module, configured to dispatch the new energy generator sets and the thermal power generator sets according to the target control strategy parameters.

[0051] Another embodiment of the present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements a power control and dispatch method based on prediction error as described in any one of the above embodiments.

[0052] Another embodiment of the present invention provides a storage medium, which includes a stored computer program. When the computer program runs, it controls the device where the storage medium is located to execute a power control and dispatch method based on prediction error as described in any one of the above embodiments.

[0053] By implementing the present invention, the following beneficial effects are achieved:

[0054] The present invention discloses a power control and dispatch method, device, terminal device, and storage medium based on prediction error. The method constructs an initial state space for characterizing the stable state of the power system by using the actual first power state value, second power state value of the new energy generator sets and thermal power generator sets in the power system at the current time period, and the prediction error state value of the new energy generator sets, so as to accurately evaluate the stable state of the power system, and constructs an initial action space according to the initial actions of each generator set regulated under the initial control strategy parameters; furthermore, by iteratively optimizing the initial control strategy parameters according to the initial state space and the initial action space, a target control strategy parameter that can eliminate the system fluctuations caused by the prediction error and restore the stability of the power system is generated, and the new energy generator sets and the thermal power generator sets are dispatched, effectively ensuring the stable operation of the power system. Description of the Drawings

[0055] Figure 1 is a flowchart of a power control and dispatch method based on prediction error provided by an embodiment of the present invention.

[0056] Figure 2 is a structural diagram of a power control and dispatch device based on prediction error provided by an embodiment of the present invention. Detailed Embodiments

[0057] To make the objectives, technical solutions, and advantages of this application clearer, the following will, in conjunction with the accompanying drawings in the embodiments of this application, clearly and completely describe the technical solutions in this application. Obviously, the described embodiments are some, rather than all, of the embodiments of this application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in this application without making creative efforts belong to the scope of protection of this application.

[0058] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which this application belongs; the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit this application; the terms "including" and "having" and any variations thereof in the description of the specification, claims, and drawings of this application are intended to cover non-exclusive inclusion.

[0059] In the description of the embodiments of this application, technical terms such as "first" and "second" are only used to distinguish different objects and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity, specific order, or primary-secondary relationship of the indicated technical features. In the description of the embodiments of this application, the meaning of "a plurality of" is more than two, unless otherwise clearly and specifically defined.

[0060] Referring to "embodiments" herein means that specific features, structures, or characteristics described in connection with the embodiments can be included in at least one embodiment of this application. The phrase appearing in various positions in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0061] In the description of the embodiments of this application, the term "and / or" is merely a description of the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this text generally represents an "or" relationship between the associated objects before and after.

[0062] In the description of the embodiments of this application, the term "a plurality of" refers to more than two (including two). Similarly, "multiple groups" refers to more than two groups (including two groups), and "multiple pieces" refers to more than two pieces (including two pieces).

[0063] In the description of the embodiments of the present application, unless otherwise clearly specified and limited, technical terms such as "installation", "connection", "connection", "fixation" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or integrated; it can also be a mechanical connection or an electrical connection; it can be directly connected, or indirectly connected through an intermediate medium, and can be the communication inside two components or the interaction relationship between two components. For those of ordinary skill in the art, the specific meanings of the above terms in the embodiments of the present application can be understood according to specific circumstances.

[0064] See Figure 1 , which is a schematic flowchart of a power control and scheduling method based on prediction error provided by an embodiment of the present invention, including:

[0065] S1. Obtain the first power state value of the new energy generating units in the power system at the current time period, the second power state value of the thermal power units at the current time period, and the predicted power state value of the new energy generating units predicted at the current time period;

[0066] In a preferred embodiment of the present invention, it is assumed that there are two types of new energy generating units, namely wind farms and photovoltaic power plants, in the power system. Therefore, the first power state value includes: the actual power state value of the wind farm at the current time period and the actual power state value of the photovoltaic power plant at the current time period; and the predicted power state value includes: the predicted power state value of the wind farm at the current time period and the predicted power state value of the photovoltaic power plant at the current time period.

[0067] Further, according to the power state values of wind, light, fire, and energy storage in the power system at the current time period a power state vector s Ω，t is represented as follows:

[0068]

[0069] where s g，t is the power state value of the thermal power unit g at time period t, is the power state value of the energy storage device s during the charging process at time period t, is the power state value of the energy storage device s during the discharging process at time period t, s w，t is the actual power state value of the wind farm w at time period t, s pv，t is the actual power state value of the photovoltaic power plant pv at time period t.

[0070] S2. Calculate the prediction error state value according to the first power state value and the predicted power state value;

[0071] In a preferred embodiment of the present invention, according to the following formula, the prediction error state values of the wind farm and the photovoltaic power are determined at the current time period,

[0072]

[0073] Among them, is the predicted power status value of wind farm w at time period t, s w，t is the actual power status value of wind farm w at time period t, is the predicted power status value of PV power station pv at time period t, s pv，t is the actual power status value of PV power station pv at time period t.

[0074] Furthermore, according to the predicted error status values of the wind farm and the photovoltaic power at the current time period, the following error status vector is constructed:

[0075]

[0076] S3. According to the first power status value, the second power status value, and the predicted error status value, construct an initial state space for characterizing the operating state of the power system, and construct an initial action space according to the change trends of the first power status value, the second power status value, and the predicted error status value under the initial control strategy parameters; wherein, the initial control strategy parameters are the control strategy parameters generated by the power system affected by the predicted error at the current time period;

[0077] Preferably, the constructing an initial state space for characterizing the operating state of the power system according to the first power status value, the second power status value, and the predicted error status value includes:

[0078] S301. Obtain the load status values of several load nodes, the power flow status values of several branches, and the charge and discharge power status values of energy storage devices in the power system;

[0079] S302. Calculate the power balance status value of the power system according to the first power status value, the second power status value, the charge and discharge power status value, and the load status value;

[0080] In a preferred embodiment of the present invention, considering that in the complex operating environment of a multi-energy system, dynamic power balance is a key factor in maintaining the stable and efficient operation of the system. Therefore, in this embodiment, the power balance status value is used as a key value in the state space. Among them, the power balance status value can comprehensively reflect the power input and output of each energy source in the system and the overall power flow situation of the system, providing a necessary basis for constructing power balance rules later, so that the formulation of the rules can closely follow the requirements of the actual operating state of the system, thereby ensuring the scientificity and effectiveness of the rules. The power balance status value is specifically expressed as follows:

[0081]

[0082] Among them, s p，t is the power balance state value, that is, the difference between the power state values of the thermal power unit, wind farm, photovoltaic power station, and energy storage device during the power generation process and the power state values of the load and the energy storage device during the charging process at time t. is the power state value of the node n where the load is located at time t.

[0083] When the power balance state value is zero, it means that the power system is in a power balance state; when the power state values of the wind, light, fire, and storage are not zero, it means that the power system is not in a power balance state.

[0084] S303. Detect the out-of-limit situation of the power state value of each branch according to the upper limit and lower limit of the power flow of several branches, and generate the upper limit detection state value and lower limit detection state value of the power flow of each branch;

[0085] In a preferred embodiment of the present invention, considering that the magnitude and direction of the line power flow reveal the power transmission situation in the power grid, the power flow action value is a key instruction for controlling and regulating the power grid, and the calculation and adjustment of the power flow action value are directly related to whether the power grid can achieve a reasonable distribution of power flow, voltage stability, and safe operation of the system; while the power flow constraint rule is one of the core rules to ensure the stable operation of the power grid. The power flow constraint rule mainly considers the physical characteristics and operation safety limitations of the power grid, and it cooperates with other rules to jointly achieve the efficient and stable operation of the power grid.

[0086] Therefore, the power state vector s transferred by the wind, light, fire, and storage to branch l at time t l，Ω，t is expressed as follows:

[0087]

[0088] Among them, s l，g，t is the power state value s of the thermal power unit g at time t g，t transferred to the power state value on branch l, s l，w，t is the power state value s of the wind farm g at time t w，t transferred to the power state value on branch l, s l，pv，t is the power state value s of the photovoltaic power station pv at time t pv，t transferred to the power state value on branch l, is the power state value during the discharge process of the energy storage device s at time t transferred to the power state value on branch l, is the power state value during the charging process of the energy storage device s at time t transferred to the power state value on branch l, is the power status value of the load at node n in time period t The power status value transferred to branch l, s l，t is the lower power limit status value of branch l in time period t, is the upper power limit status value of branch l in time period t.

[0089] Furthermore, in the complex operating environment of the power system, the reasonable distribution and stable control of power flow are crucial for the safe and efficient operation of the system. The power flow detection status values directly reflect the power transmission status in the system lines. These status values, as key indicators of the system operating status, provide indispensable data support for subsequent in-depth understanding and establishment of effective power flow constraint rules, enabling the construction of power flow constraint rules to closely fit the characteristics and requirements of power flow changes in the actual operation of the system, thus ensuring the safe and stable operation of the system. Specifically expressed as follows:

[0090]

[0091] Among them, is the upper power flow detection status value, that is, the difference between the upper power limit status value of branch l in the current time period t and the power status value of branch l in time period t, Δ s l，t is the lower power flow detection status value, that is, the difference between the power status value of branch l in the current time period t and the lower power limit status value of branch l in time period t.

[0092] When both the upper detection status value and the lower detection status value of the power flow of the wind-solar-thermal-storage lines are greater than 0, the power flow of the system branch is constrained within a reasonable range; when the upper detection status value of the power flow of the wind-solar-thermal-storage lines is less than 0 or the lower detection status value is less than 0, it represents that the power flow of branch l exceeds the limit.

[0093] S304. Construct the initial state space according to the first power status value, the second power status value, the power flow power status value, the power balance status value, the upper power flow detection status value, and the lower power flow detection status value.

[0094] In a preferred embodiment of the present invention, the expression of the initial state space is as follows:

[0095]

[0096] Preferably, constructing the initial action space according to the change trends of the first power status value, the second power status value, and the prediction error status value under the initial control strategy parameters includes:

[0097] S311. Construct a power action vector according to the first power state value, the second power state value, and the change trend of the prediction error state value under the initial control strategy parameters;

[0098] S312. When it is determined that the power system is power balanced according to the power balance state value, and it is determined that no branch of the power system has a power flow overlimit according to the power flow upper limit detection state value and the power flow lower limit detection state value, construct an initial action space according to the power action vector and the prediction error action vector.

[0099] In a preferred embodiment of the present invention, the prediction error action vector is as follows:

[0100]

[0101] is the prediction error action value of the prediction error state value of the wind farm w at time t under the policy network parameters θ (initial control strategy parameters), is the prediction error action value of the prediction error state value of the photovoltaic power station pv at time t under the policy network parameters θ.

[0102] Preferably, after constructing the prediction error action vector, it further includes:

[0103] S313. When it is determined that the power system has a power imbalance and no branch of the power system has a power flow overlimit, construct a power balance action vector according to the change trend of the power balance state value under the initial control strategy parameters, and construct an initial action space according to the power balance action vector, the power action vector, and the prediction error action vector;

[0104] S314. When it is determined that the power system is power balanced and a branch of the power system has a power flow overlimit, construct an upper limit action vector and a lower limit action vector according to the change trend of the power flow upper limit detection state value and the power flow lower limit detection state value under the initial control strategy parameters, and construct an initial action space according to the upper limit action vector, the lower limit action vector, the power action vector, and the prediction error action vector;

[0105] S315. When it is determined that there is a power imbalance in the power system and the power flow of each branch of the power system exceeds the limit, a power balance action vector is constructed according to the change trend of the power balance state value under the initial control strategy parameters, an upper limit action vector and a lower limit action vector are constructed according to the change trends of the upper limit detection state value of the power flow and the lower limit detection state value of the power flow under the initial control strategy parameters, and an initial action space is constructed according to the power balance action vector, the upper limit action vector, the lower limit action vector, the power action vector, and the prediction error action vector.

[0106] In a preferred embodiment of the present invention, when there is a power imbalance in the power system and the power flow of each branch of the power system exceeds the limit, the state space S and the action space A are expressed as follows:

[0107]

[0108] Among them are respectively the state vector and action vector of the wind-solar prediction error value, s Ω，t 、 are respectively the state vector and action vector of the wind-solar-thermal energy storage power, s l，Ω，t 、 are respectively the state vector and action vector of the wind-solar-thermal energy storage transfer power, s p，t 、a p，t are respectively the power balance state value and action value, s l，t 、 a l，t are respectively the lower detection state value and action value of the line power flow, are respectively the upper detection state value and action value of the line power flow.

[0109] Preferably, the constructing of the power balance action vector according to the change trend of the power balance state value under the initial control strategy parameters includes:

[0110] S3131. When it is indicated by the power balance state value that the power generation amount of the power system is greater than the load amount, it is determined that the first change trend constraint of the first power state value, the second power state value, and the discharge power state value of the energy storage device is negative, and the second change trend constraint of the charging power state value of the energy storage device is positive;

[0111] S3132. When it is indicated by the power balance state value that the power generation amount of the power system is less than the load amount, it is determined that the first change trend constraint is positive and the second change trend constraint is negative;

[0112] S3133. The first change trend constraint and the second change trend constraint are used to construct a power balance action constraint;

[0113] S3134. Construct a power balance action vector according to the power balance action constraint and the change trend of the power balance state value under the initial control strategy parameters.

[0114] In a preferred embodiment of the present invention, when s p，t > 0, the sum of the power generation state values of the thermal power unit, the wind farm, the photovoltaic power station, and the energy storage device is greater than the sum of the load and the energy storage device charging state value. The power generation action values of the thermal power unit, the wind farm, the photovoltaic power station, and the energy storage device are constrained in the negative domain, and the energy storage device charging action value is constrained in the positive domain.

[0115] When s p，t < 0, the sum of the power generation state values of the thermal power unit, the wind farm, the photovoltaic power station, and the energy storage device is less than the sum of the load and the energy storage device charging state value. The power generation action values of the thermal power unit, the wind farm, the photovoltaic power station, and the energy storage device are constrained in the positive domain, and the energy storage device charging action value is constrained in the negative domain. Specifically, it is expressed as follows:

[0116]

[0117]

[0118] The above formula is the interval constraint on the charge and discharge power action values of the thermal power unit, the wind farm, the photovoltaic power station, and the energy storage device when the power is unbalanced. Among them, is the interval constraint on the power action value of the thermal power unit g at time t, is the interval constraint on the power action value of the wind farm w at time t, is the interval constraint on the power action value of the photovoltaic power station pv at time t, is the interval constraint on the power action value of the energy storage device s during the discharge process at time t, is the interval constraint on the power action value of the energy storage device s during the charging process at time t, sign(s p，t ) is the judgment of the power balance state value s p，t .

[0119] After determining the constraint rules, the following power balance action vector and power balance action value are obtained:

[0120]

[0121] Among them, is the power action vector of the wind-solar-thermal-energy storage at time t, a p，t is the action value corresponding to the power balance state value s p，t .

[0122] By using the obtained power action value, the system reaches a new power balance, which is specifically expressed as follows:

[0123]

[0124] Among them, the power action values of wind, light, fire, and energy storage with power balance rules added can improve the convergence and stability of the algorithm, and enhance the response speed and dynamic performance of the system.

[0125] Preferably, constructing the upper limit action vector and the lower limit action vector according to the change trends of the upper limit detection state value of the power flow and the lower limit detection state value of the power flow under the initial control strategy parameters includes:

[0126] S3141. Determine the first branch where the power flow state value exceeds the upper limit of the power flow and the second branch where the power flow state value exceeds the lower limit of the power flow according to the upper limit detection state value of the power flow and the lower limit detection state value of the power flow;

[0127] S3142. Set the third change trend constraint of the power flow of the new energy generating unit, the thermal power unit, and the energy storage device discharging to the first branch as a negative value, and set the fourth change trend constraint of the charging power absorbed by the energy storage device from the first branch as a positive value;

[0128] S3143. Set the third change trend constraint of the power flow of the new energy generating unit, the thermal power unit, and the energy storage device discharging to the second branch as a positive value, and set the fourth change trend constraint of the charging power absorbed by the energy storage device from the second branch as a negative value;

[0129] S3144. Construct the power flow transfer constraint according to the third change trend constraint and the fourth change trend constraint of each branch;

[0130] S3145. Construct the upper limit action vector and the lower limit action vector according to the power flow transfer constraint, the change trend of the upper limit detection state value of the power flow, and the change trend of the lower limit detection state value of the power flow under the initial control strategy parameters.

[0131] In a preferred embodiment of the present invention, when Δ s l，t < 0, the power state value on branch l at time t is lower than the lower limit state value of the power on branch l at time t. The power flow state value constraints of the thermal power unit, the wind farm, the photovoltaic power station, and the energy storage device generating electricity are in the positive domain, and the power flow state value constraints of the energy storage device charging are in the negative domain; when Δ s l，tWhen >0, the power status value on branch l during time period t is higher than the power upper limit status value on branch l during time period t. The power flow status values of thermal power units, wind farms, photovoltaic power plants, and energy storage devices for power generation are constrained in the negative domain, and the power flow status values of energy storage devices for charging are constrained in the positive domain. The specific expansion is as follows:

[0132]

[0133] The above formula is the interval constraint of power flow constraints on the transfer power action values of thermal power units, wind farms, photovoltaic power plants, and energy storage devices. Among them, is the action interval constraint of the transfer power action value of thermal power unit g on branch l during time period t, is the action interval constraint of the transfer power action value of wind farm w on branch l during time period t, is the action interval constraint of the transfer power action value of photovoltaic power plant pv on branch l during time period t, is the action interval constraint of the transfer power action value of energy storage device s during the discharge process on branch l during time period t, is the action interval constraint of the transfer power action value of energy storage device s during the charging process on branch l during time period t, sign(Δ s l，t ) is the sign judgment of the detection status value Δ s l，t of the line power flow.

[0134] After determining the constraint rules, the following upper / lower limit action vector action values and upper / lower limit action vectors are obtained:

[0135]

[0136] Among them, is the power action vector of wind-solar-thermal-storage transferred to branch l during time period t, a l，t is the action value corresponding to the detection status value Δ s l，t of the line power flow, is the action value corresponding to the detection status value of the line power flow.

[0137] The power flow action value obtained through the power flow constraint rules enables the system to reach a new power flow constraint state, which is specifically expressed as follows:

[0138]

[0139] Among them, the power flow action values of wind-solar-thermal-storage with power flow constraint rules added can improve the stability of the system and the adaptability of the system to the fluctuations of renewable energy, realize the efficient utilization of energy, and improve the operation flexibility.

[0140] S4. Iteratively optimize the initial control strategy parameters according to the initial state space and the initial action space, and evaluate the stable state of the power system based on the state space updated during the iteration. When it is determined that the power system operates stably under the target control strategy parameters generated during the current iteration according to the stable state, output the target control strategy parameters;

[0141] Preferably, the step of iteratively optimizing the initial control strategy parameters according to the initial state space and the initial action space, evaluating the stable state of the power system based on the state space updated during the iteration, and outputting the target control strategy parameters when it is determined that the power system operates stably under the target control strategy parameters generated during the current iteration according to the stable state includes:

[0142] S41. Repeatedly perform parameter optimization operations according to the initial state space and the initial action space until the target control strategy parameters are generated;

[0143] S411. Obtain the strategy parameters to be optimized, the state space to be evaluated, and the action space to be evaluated. Initially, the strategy parameters to be optimized are the initial control strategy parameters, the state space to be evaluated is the initial state space, and the action space to be evaluated is the initial action space;

[0144] S412. Update the state space to be evaluated according to the action space to be evaluated, and generate an updated state space for characterizing the operating state of the power system after power dispatching according to the initial control strategy parameters;

[0145] S413. Calculate the first stable state score of the power system according to the state space to be evaluated, and calculate the second stable state score of the power system according to the updated state space;

[0146] S414. If the second stable state score does not meet the preset score threshold, use the advantage function to calculate the reward value according to the first stable state score and the second stable state score;

[0147] S415. Update the optimization trend of the strategy parameters to be optimized according to the reward value, and update the strategy parameters to be optimized according to the optimization trend to generate updated strategy parameters;

[0148] S416. Generate the action space to be evaluated required for the next round of parameter optimization operations according to the updated state space and the updated strategy parameters;

[0149] S417. Use the update policy parameter and the update state space as the policy parameter to be optimized and the state space to be evaluated required for the next round of parameter optimization operation respectively, and perform the next round of parameter optimization operation;

[0150] S418. If the second steady state score meets the preset score threshold, use the policy parameter to be optimized as the target control policy parameter.

[0151] In a preferred embodiment of the present invention, during the optimization operation, the advantage function in the F-DPPO-E algorithm provides a clear value orientation for policy evaluation by calculating the difference in rewards brought by the current action and the historical action, which helps to reduce the bias caused by inaccurate reward estimation in the traditional policy gradient algorithm, enables the algorithm to find an effective policy for the optimization operation faster, and thus accelerates the convergence process. It is specifically expressed as follows:

[0152]

[0153] where γ is the discount factor, λ is the parameter of the advantage function, and δ γ Expanded as follows:

[0154]

[0155] where, R γ (S γ ) is the reward of the advantage function for the state space S to be evaluated at the iteration number γ, and Q(S γ ) is the estimate of the value function for the state space S to be evaluated, that is, the steady state score. γ ) is the estimate of the value function for the state space S to be evaluated, that is, the steady state score. γ The update of the policy parameter in the F-DPPO-E algorithm can dynamically adjust the exploration direction of the unit start-stop strategy according to various factors such as the operating characteristics of different units, the cost function, and the system load demand, and can balance the relationship between exploring new strategies and using existing optimal strategies, which helps the algorithm to adapt to the complex constraint conditions in the optimization operation, improve the comprehensive benefit of the system, and improve the stability and convergence of the strategy.

[0156] where,

[0157]

[0158] where, represents the expected estimate of the iteration number γ, clip(ρ γ，θ , 1-ε, 1+ε) is a clipping operation that limits the value of ρ γ，θ within the range of [1-ε, 1+ε], ε is a hyperparameter used to limit the amplitude of the policy update, and ρ γ，θ is the probability ratio of the new and old policies, expanded as follows:

[0159]

[0160] where is the probability that the current policy takes action A in state S γ Take action A γ The probability of.

[0161] Finally, the Adam optimization algorithm is used to maximize the objective function L CLIP (θ), thereby updating the parameters θ of the policy network.

[0162] S5. According to the target control policy parameters, perform scheduling control on the new energy generating unit and the thermal power generating unit.

[0163] This embodiment provides a power control scheduling method based on prediction error. By constructing an initial state space for characterizing the stable state of the power system according to the actual first power state value, second power state value of the new energy generating unit and the thermal power generating unit in the current period in the power system, and the prediction error state value of the new energy generating unit, the stable state of the power system can be accurately evaluated, and an initial action space is constructed according to the initial actions of each generating unit regulated under the initial control policy parameters; furthermore, by iteratively optimizing the initial control policy parameters according to the initial state space and the initial action space, a target control policy parameter that can eliminate the system fluctuations caused by the prediction error and restore the stability of the power system is generated, and scheduling is performed on the new energy generating unit and the thermal power generating unit, effectively ensuring the stable operation of the power system.

[0164] See Figure 2 , which is a schematic structural diagram of a power control scheduling device based on prediction error provided by an embodiment of the present invention, including:

[0165] A data acquisition module, configured to acquire the first power state value of the new energy generating unit in the power system in the current period, the second power state value of the thermal power generating unit in the current period, and the predicted power state value of the new energy generating unit predicted in the current period;

[0166] An error calculation module, configured to calculate a prediction error state value according to the first power state value and the predicted power state value;

[0167] A space construction module, configured to construct an initial state space for characterizing the operating state of the power system according to the first power state value, the second power state value, and the prediction error state value, and construct an initial action space according to the change trend of the first power state value, the second power state value, and the prediction error state value under the initial control policy parameters; wherein, the initial control policy parameters are the control policy parameters generated by the power system affected by the prediction error in the current period;

[0168] A parameter optimization module, which is used to iteratively optimize the initial control strategy parameters according to the initial state space and the initial action space, evaluate the stable state of the power system according to the updated state space during the iteration process, and output the target control strategy parameters when it is determined that the power system operates stably under the target control strategy parameters generated in the current iteration process according to the stable state;

[0169] A power dispatch module, which is used to dispatch the new energy generating units and the thermal power generating units according to the target control strategy parameters.

[0170] It should be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. In addition, in the accompanying drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that they have a communication connection, which can be specifically implemented as one or more communication buses or signal lines. Those of ordinary skill in the art can understand and implement without creative efforts.

[0171] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working process of the device described above can refer to the corresponding process in the foregoing method embodiment, and will not be elaborated here.

[0172] Another preferred embodiment of the present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements a power control and dispatch method based on prediction error as described in any one of the above embodiments.

[0173] The terminal device can be a computing device such as a desktop computer, a notebook, a palm computer, and a cloud server. The terminal device may include, but is not limited to, a processor and a memory.

[0174] The so-called processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The processor is the control center of the terminal device, and connects various parts of the entire terminal device through various interfaces and lines.

[0175] The memory can be used to store the computer program. By running or executing the computer program stored in the memory and invoking the data stored in the memory, the processor realizes various functions of the terminal device. The memory mainly includes a program storage area and a data storage area. Among them, the program storage area can store an operating system, application programs required for at least one function, etc.; the data storage area can store data created according to the use of the mobile phone, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, at least one magnetic disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0176] Another preferred embodiment of the present invention provides a storage medium. The storage medium is a computer-readable storage medium, and the computer program is stored in the computer-readable storage medium. When the computer program is executed by a processor, the steps of the above-mentioned various method embodiments can be realized. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disc, a computer memory, a Read-Only Memory (ROM), a Random Access Memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0177] The above are the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements are also regarded as the protection scope of the present invention.

Claims

1. A power control and scheduling method based on prediction error, characterized in that Including: Obtain the first power status value of the new energy generating units in the power system at the current time period, the second power status value of the thermal power units at the current time period, and the predicted power status value of the new energy generating units predicted at the current time period; Calculate the prediction error status value according to the first power status value and the predicted power status value; Construct an initial state space for characterizing the operating state of the power system according to the first power status value, the second power status value, and the prediction error status value, and construct an initial action space according to the change trends of the first power status value, the second power status value, and the prediction error status value under the initial control strategy parameters; wherein, the initial control strategy parameters are the control strategy parameters generated by the power system affected by the prediction error at the current time period; Iteratively optimize the initial control strategy parameters according to the initial state space and the initial action space, evaluate the stable state of the power system according to the updated state space during the iteration process, and output the target control strategy parameters when it is determined according to the stable state that the power system operates stably under the target control strategy parameters generated during the current iteration process; Perform scheduling control on the new energy generating units and the thermal power units according to the target control strategy parameters.

2. The power control and scheduling method based on prediction error according to claim 1, wherein The constructing an initial state space for characterizing the operating state of the power system according to the first power status value, the second power status value, and the prediction error status value includes: Obtain the load status values of several load nodes in the power system, the power flow status values of several branches, and the charge and discharge power status values of energy storage devices; Calculate the power balance status value of the power system according to the first power status value, the second power status value, the charge and discharge power status value, and the load status value; Detect the over-limit conditions of the power flow status values of each branch according to the upper limit and lower limit of the power flow of several branches, and generate the upper limit detection status value and lower limit detection status value of the power flow of each branch; Construct the initial state space according to the first power status value, the second power status value, the power flow status value, the power balance status value, the upper limit detection status value of the power flow, and the lower limit detection status value of the power flow.

3. The power control and scheduling method based on prediction error according to claim 2, characterized in that The constructing an initial action space according to the change trends of the first power status value, the second power status value, and the prediction error status value under the initial control strategy parameters includes: Construct a power action vector according to the change trends of the first power status value and the second power status value under the initial control strategy parameters; Construct a prediction error action vector according to the change trend of the prediction error status value under the initial control strategy parameters; When it is determined according to the power balance status value that the power system is power balanced, and it is determined according to the upper limit detection status value of the power flow and the lower limit detection status value of the power flow that no power flow over-limit occurs in each branch of the power system, construct an initial action space according to the power action vector and the prediction error action vector.

4. The power control and scheduling method based on prediction error according to claim 3, characterized in that After constructing the prediction error action vector, it further includes: When it is determined that there is a power imbalance in the power system and no branch of the power system has a power flow limit, a power balance action vector is constructed according to the change trend of the power balance state value under the initial control strategy parameters, and an initial action space is constructed according to the power balance action vector, the power action vector, and the prediction error action vector; When it is determined that the power system is power balanced and there is a power flow limit in each branch of the power system, an upper limit action vector and a lower limit action vector are constructed according to the change trends of the upper limit detection state value and the lower limit detection state value of the power flow under the initial control strategy parameters, and an initial action space is constructed according to the upper limit action vector, the lower limit action vector, the power action vector, and the prediction error action vector; When it is determined that there is a power imbalance in the power system and there is a power flow limit in each branch of the power system, a power balance action vector is constructed according to the change trend of the power balance state value under the initial control strategy parameters, an upper limit action vector and a lower limit action vector are constructed according to the change trends of the upper limit detection state value and the lower limit detection state value of the power flow under the initial control strategy parameters, and an initial action space is constructed according to the power balance action vector, the upper limit action vector, the lower limit action vector, the power action vector, and the prediction error action vector.

5. A power control and scheduling method based on prediction error according to claim 4, characterized in that, The constructing of the power balance action vector according to the change trend of the power balance state value under the initial control strategy parameters includes: When the power generation amount of the power system represented by the power balance state value is greater than the load amount, it is determined that the first change trend constraint of the first power state value, the second power state value, and the discharge power state value of the energy storage device is negative, and the second change trend constraint of the charging power state value of the energy storage device is positive; When the power generation amount of the power system represented by the power balance state value is less than the load amount, it is determined that the first change trend constraint is positive and the second change trend constraint is negative; The first change trend constraint and the second change trend constraint are used to construct a power balance action constraint; The power balance action vector is constructed according to the power balance action constraint and the change trend of the power balance state value under the initial control strategy parameters.

6. The power control and scheduling method based on prediction error according to claim 5, wherein, The constructing of the upper limit action vector and the lower limit action vector according to the change trends of the upper limit detection state value and the lower limit detection state value of the power flow under the initial control strategy parameters includes: According to the upper limit detection state value and the lower limit detection state value of the power flow, the first branch where the power flow state value exceeds the upper power flow limit and the second branch where the power flow state value exceeds the lower power flow limit are determined; The third change trend constraint of the power flow from the new energy generating unit, the thermal power unit, and the energy storage device during discharge to the first branch is set to be negative, and the fourth change trend constraint of the charging power absorbed by the energy storage device from the first branch is set to be positive; Set the third change trend constraint of the power flow power flowing into the second branch when the new energy power generation unit, the thermal power unit, and the energy storage device discharge to a positive value, and set the fourth change trend constraint of the charging power absorbed by the energy storage device from the second branch to a negative value; Construct a power flow transfer constraint according to the third change trend constraint and the fourth change trend constraint of each branch; Construct an upper limit action vector and a lower limit action vector according to the change trends of the power flow transfer constraint, the power flow upper limit detection state value, and the power flow lower limit detection state value under the initial control strategy parameters; 7. The power control and scheduling method based on prediction error according to claim 6, characterized in that Iteratively optimize the initial control strategy parameters according to the initial state space and the initial action space, evaluate the stable state of the power system according to the updated state space during the iteration process, and output the target control strategy parameters when it is determined according to the stable state that the power system operates stably under the target control strategy parameters generated during the current iteration process, including: Repeatedly perform parameter optimization operations according to the initial state space and the initial action space until the target control strategy parameters are generated; Among them, the parameter optimization operation includes: Obtain the strategy parameters to be optimized, the state space to be evaluated, and the action space to be evaluated. Initially, the strategy parameters to be optimized are the initial control strategy parameters, the state space to be evaluated is the initial state space, and the action space to be evaluated is the initial action space; Update the state space to be evaluated according to the action space to be evaluated to generate an updated state space representing the operating state of the power system after power dispatching according to the initial control strategy parameters; Calculate the first stable state score of the power system according to the state space to be evaluated, and calculate the second stable state score of the power system according to the updated state space; If the second stable state score does not meet the preset score threshold, then use the advantage function to calculate the reward value according to the first stable state score and the second stable state score; Update the optimization trend of the strategy parameters to be optimized according to the reward value, and update the strategy parameters to be optimized according to the optimization trend to generate updated strategy parameters; Generate the action space to be evaluated required for the next round of parameter optimization operation according to the updated state space and the updated strategy parameters; Use the updated strategy parameters and the updated state space as the strategy parameters to be optimized and the state space to be evaluated required for the next round of parameter optimization operation respectively, and perform the next round of parameter optimization operation; If the second stable state score meets the preset score threshold, then use the strategy parameters to be optimized as the target control strategy parameters; 8. A power control and scheduling device based on prediction error, characterized in that, Including: A data acquisition module for acquiring the first power state value of the new energy power generation unit in the power system at the current time period, the second power state value of the thermal power unit at the current time period, and the predicted power state value of the new energy power generation unit predicted at the current time period; An error calculation module for calculating a prediction error state value according to the first power state value and the predicted power state value; A space construction module, configured to construct an initial state space for characterizing the operating state of the power system according to the first power state value, the second power state value, and the prediction error state value, and construct an initial action space according to the change trends of the first power state value, the second power state value, and the prediction error state value under the initial control strategy parameters; wherein the initial control strategy parameters are the control strategy parameters generated by the power system affected by the prediction error in the current time period. A parameter optimization module, configured to iteratively optimize the initial control strategy parameters according to the initial state space and the initial action space, evaluate the stable state of the power system according to the updated state space during the iteration process, and output the target control strategy parameters when it is determined according to the stable state that the power system operates stably under the target control strategy parameters generated in the current iteration process. A power dispatch module, configured to dispatch the new energy generating units and the thermal power generating units according to the target control strategy parameters.

9. A terminal device, characterized in that, It includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements a power control and dispatch method based on prediction error as described in any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium includes a stored computer program, wherein when the computer program runs, it controls the device where the storage medium is located to execute a power control and dispatch method based on prediction error as described in any one of claims 1 to 7.

Images