CN115642617A - Scheduling method of photovoltaic energy storage system - Google Patents
Scheduling method of photovoltaic energy storage system Download PDFInfo
- Publication number
- CN115642617A CN115642617A CN202210938501.9A CN202210938501A CN115642617A CN 115642617 A CN115642617 A CN 115642617A CN 202210938501 A CN202210938501 A CN 202210938501A CN 115642617 A CN115642617 A CN 115642617A
- Authority
- CN
- China
- Prior art keywords
- sample
- benefit
- scheduling
- state information
- power
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000004146 energy storage Methods 0.000 title claims abstract description 59
- 238000000034 method Methods 0.000 title claims abstract description 54
- 230000008901 benefit Effects 0.000 claims abstract description 114
- 238000010606 normalization Methods 0.000 claims description 4
- 230000009286 beneficial effect Effects 0.000 abstract description 3
- 230000005611 electricity Effects 0.000 description 30
- 238000010248 power generation Methods 0.000 description 17
- 238000007599 discharging Methods 0.000 description 11
- 238000010586 diagram Methods 0.000 description 10
- 238000004364 calculation method Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 230000007774 longterm Effects 0.000 description 4
- 238000011002 quantification Methods 0.000 description 2
- 230000006399 behavior Effects 0.000 description 1
- 238000012804 iterative process Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
Images
Landscapes
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention relates to the technical field of photovoltaic energy storage systems, in particular to a scheduling method of a photovoltaic energy storage system, which comprises the following steps: s1: acquiring current state information of the photovoltaic energy storage system and pre-collected standard state information; s2: acquiring a plurality of corresponding standard state information according to the current state information, and judging the current state mode; s3: generating an optimal benefit index according to the state mode and a plurality of standard state information related to the state mode; s4: and generating a scheduling parameter for controlling the photovoltaic energy storage system according to the optimal benefit index. The invention has the beneficial effects that: in the process of generating the scheduling parameters, the pre-collected standard state information is combined, so that the optimal benefit index which can be reached by the photovoltaic energy storage system in the current state mode is obtained, the scheduling parameters which accord with the current state mode are generated through the optimal benefit index and the current state information, the economic benefit maximization is achieved, and the problem that the model outputs extreme values is solved.
Description
Technical Field
The invention relates to the technical field of photovoltaic energy storage systems, in particular to a scheduling method of a photovoltaic energy storage system.
Background
A photovoltaic energy storage system refers to an equipment system which is applied to a photovoltaic power generation system and can store, convert and release circulated electric energy through an electrochemical cell or an electromagnetic energy storage medium. Because the photovoltaic power generation system is greatly influenced by external factors such as weather, seasons and the like, the phenomenon of obvious power generation power fluctuation exists in the whole working period, and further certain difficulty is brought to the dispatching of a power grid. Aiming at the problem, the output electric energy of the photovoltaic power generation system is subjected to peak regulation through the energy storage system, so that the problems of electricity abandonment, insufficient power and the like of the photovoltaic power generation system can be effectively avoided. Particularly, for an industrial park of a partial self-built photovoltaic power generation system, a local microgrid is formed by a photovoltaic power generation system, a photovoltaic energy storage system, an external power grid and power utilization equipment in the park in the industrial park usually, the local microgrid supplies power through the photovoltaic power generation system and the external power grid, redundant electric energy generated by the photovoltaic power generation system is dispatched through the photovoltaic energy storage system, and the power utilization equipment is input at a power generation valley.
In the prior art, there is a technical solution for scheduling such a scene. For example, CN201811602550.5 discloses a method and an apparatus for scheduling photovoltaic power, a computer device and a storage medium, which compare photovoltaic power generation power, photovoltaic power consumption power, a first remaining power amount, a first power consumption price and a power selling price in sequence, so as to determine how to adjust charging power of a charging pile. For another example, CN201910619256.3 discloses a multi-energy-storage-battery operation model capable of achieving smooth photovoltaic output, which is a model constructed and solved for various parameters that need to be processed during the operation process of an energy storage battery, so as to obtain the charging and discharging power of the energy storage battery.
However, in the practical implementation process, the inventor finds that, in the above technical solution, since a plurality of parameters collected on the site are simply compared through a plurality of judgment conditions, and scheduling is performed according to the comparison result, the economic benefit generated by the scheduling instruction is relatively carelessly considered in the implementation process. For example, in the above scheme, it is stated that "when the second electricity price is less than or equal to the electricity price for buying, the photovoltaic charging power of the charging pile is reduced according to the photovoltaic power generation power and the photovoltaic power of the energy storage system in the current time period", which is adjusted only for the photovoltaic charging power at the current time, and the economic benefit cannot be maximized in the relatively long-term scheduling process of the photovoltaic energy storage system. For another example, only a single model is used for planning the scheduling parameters, and extreme values which do not conform to real environment conditions are generated due to strong target guidance, so that the photovoltaic energy storage system cannot be effectively controlled.
Disclosure of Invention
Aiming at the problems in the prior art, a scheduling method of a photovoltaic energy storage system is provided.
The specific technical scheme is as follows:
a scheduling method of a photovoltaic energy storage system comprises the following steps:
step S1: acquiring current state information of the photovoltaic energy storage system and a plurality of different pre-collected standard state information;
step S2: acquiring a plurality of corresponding standard state information according to the current state information, and judging the current state mode of the photovoltaic energy storage system;
and step S3: generating an optimal benefit indicator corresponding to the status pattern according to the status pattern and a plurality of standard status information associated with the status pattern;
and step S4: and generating a scheduling parameter for controlling the photovoltaic energy storage system according to the optimal benefit index.
Preferably, the current status information includes a plurality of status parameters, and the step S2 includes:
step S21: sequentially comparing a plurality of state parameters in the current state information according to a preset comparison rule, and generating corresponding pre-scheduling information according to a comparison result;
step S22: and determining the current state mode according to the pre-scheduling information and the standard state information.
Preferably, the step S22 includes:
step S221: respectively generating a first sample distance between a sample point of the pre-scheduling information and a sample point of each piece of standard state information;
step S222: clustering the sample points according to the first sample distance to generate a plurality of sample classes;
step S223: respectively calculating a second sample distance between every two sample classes, and clustering the sample classes according to the second sample distance;
step S224: and repeating the step S223 until the number of the clustered sample classes reaches a set class threshold, and outputting the sample class in which the sample point of the pre-scheduling information is located as the state pattern.
Preferably, the step S3 includes:
step S31: acquiring a plurality of standard state information under the state mode, and respectively calculating the benefit index of each standard state information;
step S32: and generating the optimal benefit index in the state mode according to the plurality of benefit indexes.
Preferably, in step S31, the method for calculating the benefit index includes:
wherein ,Rhat In order to be the benefit index,
for the power on grid for the ith hour,
is the price of the power on line in the ith hour,
for the lower grid power of the ith hour,
and (5) supplying power to the power grid at the ith hour.
Preferably, in step S32, the method for calculating the optimal benefit index includes:
wherein ,
as the optimum benefit index, R k Is the number, R, of all the standard status information in the status mode hat ∈top10％R k Is the standard status information, R, with the benefit indicator being the top 10% in the status mode hat The benefit indicator is the standard status information.
Preferably, the step S4 includes:
step S41: generating a sample to be predicted according to the pre-standard state information and the optimal benefit index, and generating a benefit sample according to the standard state information and the benefit index corresponding to the standard state information;
step S42: respectively calculating a second sample distance between the sample point of the sample to be predicted and the sample point of each benefit sample;
step S43: generating a benefit sample set corresponding to the sample to be predicted according to the second sample distance;
step S44: and generating the scheduling parameters corresponding to the samples to be predicted according to the benefit sample set.
Preferably, before the step S42, normalization processing is further performed on all the state parameters in the sample to be predicted and the benefit sample, respectively;
wherein ,xnew The normalized state parameters are represented by X, max (X) is the maximum value of the state parameters in the same class, and min (X) is the minimum value of the state parameters in the same class.
Preferably, in step S44, the method for generating the scheduling parameter according to the standard status information set includes:
wherein ,Pbat* For the scheduling parameter corresponding to the sample to be predicted, k is the number of the benefit samples in the benefit sample set, P bat For the benefit sample, S, in the benefit sample set g For the benefit sample set, P bat Scheduling parameters for the benefit samples.
The technical scheme has the following advantages or beneficial effects: in the process of generating the scheduling parameters, the pre-acquired standard state information is combined, so that the optimal benefit index which can be reached by the photovoltaic energy storage system in the current state mode is obtained, the scheduling parameters which accord with the current state mode are generated through the optimal benefit index and the current state information, the economic benefit maximization is achieved, and the problem that the model outputs an extreme value is avoided.
Drawings
Embodiments of the present invention will now be described more fully hereinafter with reference to the accompanying drawings. The drawings are, however, to be regarded as illustrative and explanatory only and are not restrictive of the scope of the invention.
FIG. 1 is an overall schematic diagram of an embodiment of the present invention;
FIG. 2 is a schematic view of a photovoltaic energy storage system in an embodiment of the invention;
FIG. 3 is a diagram illustrating the substep of step S2 in the embodiment of the present invention;
FIG. 4a is a diagram illustrating the substep of step S21 according to an embodiment of the present invention;
FIG. 4b is a diagram illustrating the substep of step S21 according to an embodiment of the present invention;
FIG. 5a is a schematic diagram illustrating the substep of step S21 according to another embodiment of the present invention;
FIG. 5b is a schematic diagram illustrating the substep of step S21 according to another embodiment of the present invention;
FIG. 6 is a diagram illustrating the substep of step S22 according to an embodiment of the present invention;
FIG. 7 is a schematic diagram illustrating the substep of step S3 in the embodiment of the present invention;
FIG. 8 is a diagram illustrating the substep of step S4 in the embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It should be noted that the embodiments and features of the embodiments may be combined with each other without conflict.
The invention is further described with reference to the following drawings and specific examples, which are not intended to be limiting.
The invention includes:
a method for scheduling a photovoltaic energy storage system, as shown in fig. 1, includes:
step S1: acquiring current state information of the photovoltaic energy storage system and a plurality of different pre-collected standard state information;
step S2: acquiring a plurality of corresponding standard state information according to the current state information, and judging the current state mode of the photovoltaic energy storage system;
and step S3: generating an optimal benefit index corresponding to the state mode according to the state mode and a plurality of standard state information related to the state mode;
and step S4: and generating a scheduling parameter for controlling the photovoltaic energy storage system according to the optimal benefit index.
Specifically, aiming at the problem that when the photovoltaic energy storage system in the prior art is scheduled, corresponding scheduling instructions are obtained only by simply judging according to a plurality of state parameters, and the economic benefit maximization is difficult to realize in the long-term operation process, in the embodiment, a plurality of different standard state information are collected in advance before the scheduling is started, and the standard state information respectively corresponds to different states of the photovoltaic energy storage system during the operation. When the dispatching is started, the current state information at the current moment is obtained, clustering is carried out according to the current state information and the standard state information, so that the state mode of the microgrid where the photovoltaic energy storage system is located is obtained, the optimal benefit index in the state mode is generated, the dispatching parameter can be obtained through regression of the optimal benefit index, the photovoltaic energy storage system is controlled to carry out charging and discharging operations through the dispatching parameter, and the maximization of economic benefit is achieved.
Further, for the photovoltaic energy storage system in the prior art, when the linear programming model is used for scheduling, the problem of extreme values which cannot be realized due to the guidance of a strong objective function is easily output.
In the implementation process, the scheduling method is arranged in the photovoltaic energy storage system as a software embodiment. As shown in fig. 2, the photovoltaic energy storage system 101 is disposed in a microgrid 102, the microgrid 102 refers to a local microgrid including a photovoltaic power generation system 102 and a power utilization system 104, the local microgrid is further connected to an external power grid 105, and when a partial condition is met, such as photovoltaic power generation power or energy storage scheduling power is insufficient to meet a load of the power utilization system 104, power is taken from the external power grid 105, and when a specific condition is met, power is transmitted to an external power grid A5 to obtain an online electric charge. The current state information refers to state parameters acquired from the photovoltaic energy storage system in real time and comprises battery residual capacity SOC and photovoltaic power P pv Load power P load External network electricity price C out Photovoltaic internet price C pv Wait stateAnd the set of state parameters can represent the current running state of the photovoltaic energy storage system. Standard status information refers to historical scheduling data prior to the current start of scheduling, which is collected accordingly&The storage program is connected to the photovoltaic energy storage system and stores the state parameters of different periods.
In a preferred embodiment, as shown in fig. 3, step S2 comprises:
step S21: sequentially comparing a plurality of state parameters in the current state information according to a preset comparison rule, and generating corresponding pre-scheduling information according to a comparison result;
step S22: and determining the current state mode according to the pre-scheduling information and the standard state information.
Specifically, in the embodiment, after the pre-scheduling information is generated by comparing the state parameters, the pre-scheduling information is processed according to the pre-acquired standard state information and the pre-scheduling information, so that the state mode of the current state information in the overall scheduling process is judged on the basis of the long-term accumulated standard state information, and then the optimal benefit index corresponding to the current state mode can be generated according to different state modes in the subsequent judging process, so that the scheduling parameter obtained by regression of the optimal benefit index is the scheduling parameter which can be realized in the state mode, and the problem that the model output deviates from the actual extreme value is avoided.
In the implementation process, the pre-scheduling information refers to pre-scheduling information corresponding to current state information generated by sequentially comparing the state parameters with preset judgment conditions, and in this embodiment, the pre-scheduling information is only used for representing the current operating state of the photovoltaic energy storage system and is not equal to a control instruction or a scheduling parameter actually used for the photovoltaic energy storage system. In general, as shown in table 1, the standard status information and the pre-scheduling information each have the following fields: battery remaining capacity SOC, photovoltaic power P pv Load power P load Outside, inGrid price C out Photovoltaic internet price C pv Charge and discharge power p of energy storage battery bat Power over grid power p up Lower grid power p down . Wherein, the battery residual capacity SOC and the photovoltaic power P pv Load power P load Outside network electricity price C out Photovoltaic internet price C pv A field in the category of "operation mode", which is derived from the state parameters received by the photovoltaic energy storage system during operation; charging and discharging power p of energy storage battery bat Power on grid power p up Lower grid power p down And the field is a field of a control instruction category, and is used for representing the scheduling condition of the photovoltaic energy storage system in a certain operation mode. Wherein, the charging and discharging power p of the energy storage battery bat And when the field is a negative number, the photovoltaic energy storage system is in a charging state. It should be noted that the above mentioned convention of sign only indicates the setting in one embodiment, and the opposite can be set according to actual needs.
TABLE 1
Based on the above parameters, we can use several fixed parameters collected in advance, such as the maximum battery capacity SOC _ max, the minimum battery capacity SOC _ min, the maximum battery discharge power P _ bat _ out _ max, the maximum battery charge power P _ bat _ in _ max, and various determination conditions, such as "whether the photovoltaic power per unit time is greater than the load power", "whether the battery residual capacity per unit time is less than the maximum battery capacity"; "whether the photovoltaic grid-surfing price is greater than the unit time external grid electricity price"; and step-by-step comparison is carried out by taking the maximum charging power in the unit time of the battery as a difference value between the photovoltaic power and the load power in the unit time and the like as comparison rules, so that the approximate running state of the current microgrid is judged, and corresponding scheduling parameters are generated. The multiple judgment conditions do not have strict execution sequence among themselves.
For example, in an embodiment, as shown in fig. 4a and 4b, the step S21 of generating the pre-scheduling information by comparing the rules includes:
step A21: judging whether the photovoltaic power in unit time is greater than the load power;
if yes, go to step A22
If not, turning to the step A25:
step A22: judging whether the residual electric quantity of the battery in unit time is smaller than the maximum capacity of the battery;
if yes, go to step a23:
if not, setting the scheduling parameter in the pre-scheduling information as 'selling electricity to an external power grid according to the difference value of the photovoltaic power and the load power in unit time';
step A23: judging whether the photovoltaic internet price is larger than the unit time external grid electricity price or not;
if yes, setting the scheduling parameters in the pre-scheduling information as 'selling electricity to an external power grid according to the difference value of the photovoltaic power and the load power in unit time';
if not, turning to the step A24;
step A24: judging whether the maximum charging power of the battery in unit time is larger than the difference value between the photovoltaic power and the load power in unit time;
if yes, setting the scheduling parameter in the pre-scheduling information as 'charging the battery by the difference value of the photovoltaic power and the load power in unit time';
if not, setting the scheduling parameter in the pre-scheduling information as 'charging the battery with the maximum charging power in the unit time of the battery, and selling electricity to an external power grid according to the residual power';
step A25: judging whether the residual electric quantity of the battery in unit time is larger than the minimum capacity of the battery;
if yes, go to step A26;
if not, turning to the step A29;
step A26: judging whether the photovoltaic internet price is larger than the unit time external grid electricity price or not;
if yes, go to step A27;
if not, turning to the step A28;
step A27: judging whether the residual electric quantity of the battery in unit time is smaller than the maximum capacity of the battery;
if yes, setting the scheduling parameters in the pre-scheduling information as 'charging the battery with the maximum charging power of the battery in unit time, and simultaneously getting electricity to an external power grid according to the difference value of the load power and the photovoltaic power in unit time and the charging rate of the battery';
if not, setting the scheduling parameter in the pre-scheduling information as 'getting electricity to an external power grid according to the difference value of the load power and the photovoltaic power in unit time';
step A28: judging whether the maximum discharge power of the battery in unit time is greater than the load power in unit time;
if yes, setting the scheduling parameter in the pre-scheduling information as 'the battery discharges with the load power';
if not, setting the scheduling parameter in the pre-scheduling information as 'discharging the battery with the maximum discharging power, and simultaneously getting electricity to an external power grid according to the difference value of the load power, the battery discharging power and the photovoltaic power generation power';
step A29: judging whether the photovoltaic internet price is larger than the unit time external grid electricity price or not;
if yes, setting the scheduling parameter in the pre-scheduling information as 'charging the battery with the maximum charging power in unit time of the battery, and simultaneously getting electricity to an external power grid according to the difference value of the load power and the photovoltaic power in unit time and the charging rate of the battery';
if not, setting the scheduling parameter in the pre-scheduling information as 'getting electricity to an external power grid according to the difference value of the load power and the photovoltaic power in unit time'.
Alternatively, in another embodiment, as shown in fig. 5a and fig. 5b, in step S21, the process of generating the pre-scheduling information by comparing the rules includes:
step B20: judging whether the real-time electricity price of the external power grid is less than or equal to the electricity price when the external power grid is flat (flat peak valley);
if yes, go to step B21;
if not, turning to the step B24:
step B21: judging whether the residual electric quantity of the battery in unit time is smaller than the maximum capacity of the battery or not;
if yes, setting the scheduling parameter in the pre-scheduling information as 'charging the battery with the maximum charging power in the unit time of the battery';
if not, the step B22 is switched to:
step B22: judging whether the photovoltaic power in unit time is greater than the load power;
if yes, setting the scheduling parameters in the pre-scheduling information as 'selling electricity to an external power grid according to the difference value of the photovoltaic power and the load power in unit time';
if not, turning to step B23;
step B23: judging whether the maximum discharge power of the battery in unit time is smaller than the difference value between the load power and the photovoltaic power in unit time;
if yes, setting the scheduling parameters in the pre-scheduling information as 'discharging the battery with the maximum discharging power, and simultaneously taking electricity to an external power grid according to the difference value of the load power, the battery discharging power and the photovoltaic power generation power';
if not, setting the scheduling parameter in the pre-scheduling information as 'the battery discharges by the difference value of the load power and the photovoltaic power in unit time';
step B24: judging whether the photovoltaic power in unit time is greater than the load power;
if yes, go to step B25;
if not, turning to step B27;
step B25: judging whether the residual electric quantity of the battery in unit time is smaller than the maximum capacity of the battery;
if yes, go to step B26;
if not, setting the scheduling parameter in the pre-scheduling information as 'selling electricity to an external power grid according to the difference value of the photovoltaic power and the load power in unit time';
step B26: judging whether the maximum charging power of the battery in unit time is smaller than the difference value between the photovoltaic power and the load power in unit time;
if yes, setting the scheduling parameters in the pre-scheduling information as 'charging the battery with the maximum charging power in the unit time of the battery, and selling electricity to an external power grid according to the residual power';
if not, setting the scheduling parameter in the pre-scheduling information as 'charging the battery according to the difference value of the photovoltaic power and the load power in unit time';
step B27: judging whether the residual electric quantity of the battery in unit time is larger than the minimum capacity of the battery;
if yes, go to step B28;
if not, setting the scheduling parameter in the pre-scheduling information as 'getting electricity to an external power grid according to the difference value of the load power and the photovoltaic power in unit time';
step B28: judging whether the maximum discharge power of the battery in unit time is smaller than the difference value between the load power and the photovoltaic power in unit time;
if yes, setting the scheduling parameter in the pre-scheduling information as 'the battery is discharged by the difference value of the load power and the photovoltaic power in unit time';
if not, setting the scheduling parameter in the pre-scheduling information as 'the battery discharges with the maximum discharge power, and simultaneously getting electricity to an external power grid according to the difference value of the load power, the battery discharge power and the photovoltaic power generation power'.
It will be apparent that in some alternative implementations, the functions noted in the flowcharts or block diagrams may occur out of the order noted in the figures. For example, two steps or two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved.
In a preferred embodiment, as shown in fig. 6, step S22 comprises:
step S221: respectively generating a first sample distance of a sample point of the pre-scheduling information and a sample point of each piece of standard state information;
step S222: clustering the sample points according to the first sample distance to generate a plurality of sample classes;
step S223: respectively calculating a second sample distance between every two sample classes, and clustering the sample classes according to the second sample distance;
step S224: and step S223 is repeated until the number of the clustered sample classes reaches the set class threshold, and the sample class where the sample point of the pre-scheduling information is located is output as the state mode.
Specifically, in order to realize better division of the state mode to which the current state information belongs, a KNN method is introduced in the embodiment to classify the sample points of the pre-scheduling information and the standard state information, and the set of the plurality of sample points is converged into a plurality of specific state modes through the processes of clustering, calculating the centroid and clustering again. The number of the state modes is selected according to the number of the collected standard state information in the actual operation process, and the state modes represent the operation state of a micro-grid in which the photovoltaic energy storage system is positioned, such as high battery residual capacity and high online electricity price; the states of low load power, low off-grid electricity price and the like. By clustering the pre-scheduling information and the standard state information and dividing the pre-scheduling information and the standard state information into a certain operation mode, the optimal benefit index which can be reached in the certain operation mode can be judged in the subsequent process according to the pre-collected standard state information, and then the scheduling parameter which should be executed by the current state information is reversely deduced according to the optimal benefit index and the standard state information, so that the overall optimization of the operation mode of the photovoltaic energy storage system is realized. It should be noted that, while the number of the state patterns is selected, it is necessary to consider that too many state patterns occupy too many computing resources in the subsequent regression operation process, which results in a prolonged processing time.
In practice, the above process is embodied as an iterative process of calculating distances, clustering based on distances, calculating centroids, and then calculating distances again. In the first calculation, step S222, a distance calculation method common to the KNN method is adopted, that is, according to the fields in the pre-schedule information and the standard state information, including the above-mentioned battery remaining capacity SOC and the photovoltaic power P pv Load power P load External network electricity price C out Photovoltaic internet price C pv And the fields are used for respectively calculating the distance between each sample point.The distance used here is a euclidean distance, such as:
wherein d (i, j) indicates the distance between the ith and jth sample points, i m M field for i sample point, j m The number of the m-th field of the jth sample point is n, which is 5 in this embodiment.
After the calculation of the first sample distance of each sample point is completed, clustering can be performed through a pre-divided distance threshold, and a certain pair of sample points with the first sample distance smaller than the distance threshold is taken as a class, so that a first clustering process is completed. And then, calculating the centroid of the class as a class center for each clustered sample class, calculating the distance between the class centers again as a sample class distance, judging the calculated sample class distance by adopting a sample distance threshold, and combining a certain pair of sample classes of which the sample class distance is smaller than the sample distance threshold to form a new sample class. By repeating the clustering process of the sample classes, a plurality of sample classes can be converged into a plurality of sample classes with specific number, namely a plurality of state patterns, so that the state patterns of the pre-scheduling information can be judged.
The class center calculation process of the sample class is as follows:
wherein ,ci Is a sample class R i S is a sample class R i The coordinates of each sample point.
As an alternative embodiment, the category threshold is set to 4.
In a preferred embodiment, as shown in fig. 7, step S3 comprises:
step S31: acquiring a plurality of standard state information under a state mode, and respectively calculating the benefit index of each standard state information;
step S32: and generating the optimal benefit index in the state mode according to the plurality of benefit indexes.
In a preferred embodiment, in step S31, the method for calculating the benefit index includes:
wherein ,Rhat Is used as an index of the benefit,
for the power on grid for the ith hour,
is the price of the power on line in the ith hour,
for the lower grid power of the ith hour,
and (4) supplying power to the power grid at the ith hour.
Specifically, in order to achieve better quantification of economic benefits generated by the charging and discharging behaviors of the photovoltaic energy storage system, in this embodiment, the benefit index is set to be 1/24 of the difference between the photovoltaic electricity selling income ending 24 hours a day and the outward electricity purchasing cost, so that a better quantification effect is achieved.
In a preferred embodiment, in step S32, the method for calculating the optimal benefit index includes:
wherein ,
for the optimum benefit index, R k Is a state modelNumber of all standard state information under the formula, R hat ∈top10％R k Is the standard status information with the benefit index of the top 10% in the status mode, R hat Is the benefit index of the standard state information.
Specifically, for the photovoltaic energy storage system in the prior art, when the linear programming model is used for scheduling, the problem of the extreme value that cannot be realized due to the guidance of the strong objective function is easily output, in this embodiment, after the benefit index of each piece of standard state information in the state mode is calculated, all pieces of standard state information in the state mode are sorted from large to small according to the benefit index, and the average value of the benefit indexes is obtained by extracting the standard state information of 10% of the first order to serve as the optimal benefit index in the state mode, so that the problem of the extreme value output by the model is avoided.
In a preferred embodiment, as shown in fig. 8, step S4 comprises:
step S41: generating a sample to be predicted according to the pre-standard state information and the optimal benefit index, and generating a benefit sample according to the standard state information and the benefit index corresponding to the standard state information;
step S42: respectively calculating a second sample distance between the sample point of the sample to be predicted and the sample point of each benefit sample;
step S43: generating a benefit sample set corresponding to the sample to be predicted according to the second sample distance;
step S44: and generating a scheduling parameter corresponding to the sample to be predicted according to the benefit sample set.
Specifically, for the problem that when the photovoltaic energy storage system in the prior art is scheduled, it is difficult to maximize the economic benefit in the long-term operation process only by simply determining according to a plurality of state parameters to obtain a corresponding scheduling instruction, in this embodiment, after calculating the optimal benefit index, the optimal benefit index is combined with the pre-scheduling information to generate a sample to be predicted, and a benefit sample is generated according to the standard state information and the benefit index corresponding to the standard state information, as shown in table 2 below:
TABLE 2
It should be noted that, the to-be-predicted sample generated here has a field of "control instruction" portion removed with respect to the pre-scheduling information, and a field of "benefit indicator" added, and the benefit sample has fields of "operation mode", "control instruction" and "benefit indicator" at the same time. By generating the sample to be predicted and the benefit sample, the scheduling parameter corresponding to the current state information, namely the relevant parameter of the control instruction part, can be obtained through regression according to the optimal benefit index and the current state information in the process, so that the optimal benefit index in the current mode is realized.
In order to implement the above process, in this embodiment, a plurality of benefit sample sets with relatively close to-be-predicted point locations are obtained by calculating the second sample distance, and an average value is obtained according to a "control instruction" in the benefit sample sets, so that effective prediction of a scheduling parameter that should be implemented by a to-be-predicted sample is implemented.
In a preferred embodiment, before step S42, normalization processing is further performed on all the state parameters in the sample to be predicted and the benefit sample, respectively;
wherein ,xnew The normalized state parameter is X, the normalized state parameter is max (X), the maximum value of the same type of state parameter is max (X), and the minimum value of the same type of state parameter is min (X).
Specifically, in order to achieve a better prediction effect on the scheduling parameters, in this embodiment, before the distance between the sample to be predicted and the benefit sample is calculated as the second sample, normalization processing is performed on each state parameter to generate a state parameter that is standardized and has a unified dimension, so as to achieve a better prediction effect.
In a preferred embodiment, the method for generating the scheduling parameter according to the standard status information set in step S44 includes:
wherein ,pbat* For the scheduling parameter corresponding to the sample to be predicted, k is the number of benefit samples in the set of standard state information, p bat For benefit samples in the set of standard status information, S g Is a standard set of state information, P bat Scheduling parameters for benefit samples.
Specifically, in order to achieve a better prediction effect of the scheduling parameter of the sample to be predicted, in this embodiment, the benefit samples near the sample to be predicted are clustered based on the second sample distance in the prediction process, and the mean value of the class is obtained through calculation, so as to generate the scheduling parameter, and further make the scheduling parameter conform to the optimal benefit index in the current state mode.
The invention has the beneficial effects that: in the process of generating the scheduling parameters, the optimal benefit indexes which can be realized in different state modes are calculated, and the scheduling parameters which should be executed by the current state information are predicted by combining the optimal benefit indexes and the standard state information, so that the scheduling parameters which are finally used for controlling the photovoltaic energy storage system can realize the benefit maximization. Meanwhile, the current state information and the pre-collected standard state information are clustered, and the state mode to which the current state information belongs is judged to generate the optimal benefit index which can be realized under different state modes, so that the problem that an extreme scheduling instruction which does not accord with the actual condition and cannot be implemented is easily generated by predicting through a fixed single model in the prior art is avoided, and the interpretability and the conservation of the output instruction are ensured.
While the invention has been described with reference to a preferred embodiment, it will be understood by those skilled in the art that various changes in form and detail may be made without departing from the spirit and scope of the invention.
Claims (9)
1. A scheduling method of a photovoltaic energy storage system is characterized by comprising the following steps:
step S1: acquiring current state information of the photovoltaic energy storage system and a plurality of different pre-collected standard state information;
step S2: acquiring a plurality of corresponding standard state information according to the current state information, and judging the current state mode of the photovoltaic energy storage system;
and step S3: generating an optimal benefit indicator corresponding to the status pattern according to the status pattern and a plurality of standard status information associated with the status pattern;
and step S4: and generating a scheduling parameter for controlling the photovoltaic energy storage system according to the optimal benefit index.
2. The scheduling method of claim 1, wherein the current status information comprises a plurality of status parameters, and the step S2 comprises:
step S21: sequentially comparing a plurality of state parameters in the current state information according to a preset comparison rule, and generating corresponding pre-scheduling information according to a comparison result;
step S22: and determining the current state mode according to the pre-scheduling information and the standard state information.
3. The scheduling method according to claim 2, wherein the step S22 comprises:
step S221: respectively generating a first sample distance between a sample point of the pre-scheduling information and a sample point of each piece of standard state information;
step S222: clustering the sample points according to the first sample distance to generate a plurality of sample classes;
step S223: respectively calculating a second sample distance between every two sample classes, and clustering the sample classes according to the second sample distance;
step S224: and repeating the step S223 until the number of the clustered sample classes reaches a set class threshold, and outputting the sample class in which the sample point of the pre-scheduling information is located as the state pattern.
4. The scheduling method according to claim 1, wherein the step S3 comprises:
step S31: acquiring a plurality of standard state information under the state mode, and respectively calculating the benefit index of each standard state information;
step S32: and generating the optimal benefit index in the state mode according to the plurality of benefit indexes.
5. The scheduling method of claim 4, wherein in the step S31, the calculating method of the benefit index comprises:
wherein ,Rhat In order to be the benefit index,
for the power on grid for the ith hour,
is the price of the power on line in the ith hour,
for the lower grid power of the ith hour,
and (4) supplying power to the power grid at the ith hour.
6. The scheduling method of claim 4, wherein in the step S32, the calculating method of the optimal benefit index comprises:
wherein ,
as the optimum benefit index, R k Is the number, R, of all the standard status information in the status mode hat ∈top10％R k Is the standard state information, R, of which the benefit index is the top 10% in the state mode hat The benefit indicator is the standard status information.
7. The scheduling method according to claim 2, wherein the step S4 comprises:
step S41: generating a sample to be predicted according to the pre-standard state information and the optimal benefit index, and generating a benefit sample according to the standard state information and the benefit index corresponding to the standard state information;
step S42: respectively calculating a second sample distance between the sample point of the sample to be predicted and the sample point of each benefit sample;
step S43: generating a benefit sample set corresponding to the sample to be predicted according to the second sample distance;
step S44: generating the scheduling parameters corresponding to the samples to be predicted according to the benefit sample set.
8. The scheduling method according to claim 7, wherein before the step S42, normalization processing is further performed on all the state parameters in the to-be-predicted sample and the benefit sample, respectively;
wherein ,xnew The normalized state parameter is X, max (X) is the maximum value of the state parameters of the same class, and min (X) is the minimum value of the state parameters of the same class.
9. The scheduling method of claim 7, wherein in step S44, the method for generating the scheduling parameter according to the standard status information set comprises:
wherein ,Pbat* For the scheduling parameter corresponding to the sample to be predicted, k is the number of the benefit samples in the benefit sample set, P bat For the benefit samples, S, in the benefit sample set g For the benefit sample set, P bat Scheduling parameters for the benefit samples.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210938501.9A CN115642617B (en) | 2022-08-05 | 2022-08-05 | Scheduling method of photovoltaic energy storage system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210938501.9A CN115642617B (en) | 2022-08-05 | 2022-08-05 | Scheduling method of photovoltaic energy storage system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115642617A true CN115642617A (en) | 2023-01-24 |
| CN115642617B CN115642617B (en) | 2023-06-09 |
Family
ID=84940488
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210938501.9A Active CN115642617B (en) | 2022-08-05 | 2022-08-05 | Scheduling method of photovoltaic energy storage system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115642617B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007295680A (en) * | 2006-04-24 | 2007-11-08 | Matsushita Electric Ind Co Ltd | Load control device |
| CN107153892A (en) * | 2017-05-12 | 2017-09-12 | 沈阳工程学院 | Distributed wind and optical energy integrated generated power forecasting system |
| CN114421530A (en) * | 2022-02-17 | 2022-04-29 | 科大智能科技股份有限公司 | Scheduling method and device of photovoltaic energy storage system |
| US20220140610A1 (en) * | 2020-10-30 | 2022-05-05 | Hygge Energy Inc. | Methods and systems for green energy charging of electrical vehicles |
-
2022
- 2022-08-05 CN CN202210938501.9A patent/CN115642617B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007295680A (en) * | 2006-04-24 | 2007-11-08 | Matsushita Electric Ind Co Ltd | Load control device |
| CN107153892A (en) * | 2017-05-12 | 2017-09-12 | 沈阳工程学院 | Distributed wind and optical energy integrated generated power forecasting system |
| US20220140610A1 (en) * | 2020-10-30 | 2022-05-05 | Hygge Energy Inc. | Methods and systems for green energy charging of electrical vehicles |
| CN114421530A (en) * | 2022-02-17 | 2022-04-29 | 科大智能科技股份有限公司 | Scheduling method and device of photovoltaic energy storage system |
Non-Patent Citations (1)
| Title |
|---|
| 张海涛: "微电网并网运行优化调度策略及应用", 《中国优秀硕士学位论文全文数据库》, no. 1, pages 1 - 72 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115642617B (en) | 2023-06-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111062632B (en) | 5G energy Internet virtual power plant economic dispatching method based on edge intelligence | |
| CN112687969B (en) | Synchronous retirement cooperative control method, system and medium based on energy storage system | |
| CN111555347A (en) | Wind power prediction-based wind storage combined operation control method | |
| CN113022361A (en) | Charging pile-based ordered charging method under power demand response condition | |
| CN109572478A (en) | Charging station electric car steady charging system and method online | |
| CN117254464B (en) | Control method and system of energy storage system | |
| CN115733178A (en) | Optical storage charging station capacity configuration method and system based on cost and risk multiple targets | |
| CN114358378A (en) | User side energy storage optimal configuration system and method for considering demand management | |
| CN105470947B (en) | A kind of micro-capacitance sensor dispatching method based on quantum-behaved particle swarm optimization | |
| CN110707711A (en) | User-side comprehensive energy system hierarchical regulation and control method and system | |
| CN117172486A (en) | Reinforced learning-based virtual power plant optical storage resource aggregation regulation and control method | |
| CN115642617B (en) | Scheduling method of photovoltaic energy storage system | |
| CN113283654B (en) | Micro-grid group optimization control method and system based on aggregation and disconnection mechanism | |
| CN115238992A (en) | Power system source load storage coordination optimization method and device and electronic equipment | |
| CN115622096A (en) | Load characteristic-based capacity configuration method and system for grid-connected micro-grid system | |
| CN109149644B (en) | Light-storage integrated online strategy matching and collaborative optimization method based on big data analysis | |
| CN114447963A (en) | Energy storage battery power control method and system | |
| CN112821451A (en) | Distributed photovoltaic access coping method for smart town power distribution network based on demand side management and energy storage | |
| CN112613654A (en) | Comprehensive energy system flexibility evaluation method based on multi-type energy storage | |
| Meng et al. | Capacity optimization configuration of hybrid energy storage system using a modified grey wolf optimization | |
| CN112766733B (en) | Method and system for accelerating convergence of optimized scheduling algorithm by using improved K-means algorithm | |
| Yu et al. | GIS-based optimal siting and sizing of substation using semi-supervised learning | |
| CN113283755B (en) | Intelligent scheduling decision-making method for workshops | |
| Su et al. | Aggregation-based Self-dispatch of the Virtual Power Plant with Massive Controllable Batteries | |
| CN117424261A (en) | Light storage and charge cooperative control heavy overload treatment method and system based on energy router |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |