CN117674301B - Comprehensive energy control method based on neural network - Google Patents
Comprehensive energy control method based on neural network Download PDFInfo
- Publication number
- CN117674301B CN117674301B CN202410129667.5A CN202410129667A CN117674301B CN 117674301 B CN117674301 B CN 117674301B CN 202410129667 A CN202410129667 A CN 202410129667A CN 117674301 B CN117674301 B CN 117674301B
- Authority
- CN
- China
- Prior art keywords
- energy storage
- energy
- park
- curve
- estimated
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 29
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 19
- 238000004146 energy storage Methods 0.000 claims abstract description 235
- 230000008859 change Effects 0.000 claims abstract description 21
- 238000005457 optimization Methods 0.000 claims abstract description 15
- 238000010248 power generation Methods 0.000 claims abstract description 15
- 238000005265 energy consumption Methods 0.000 claims description 69
- 238000012937 correction Methods 0.000 claims description 49
- 238000007781 pre-processing Methods 0.000 claims description 26
- 238000012216 screening Methods 0.000 claims description 18
- 238000012545 processing Methods 0.000 claims description 17
- 238000006243 chemical reaction Methods 0.000 claims description 16
- 230000005611 electricity Effects 0.000 claims description 8
- 230000008569 process Effects 0.000 claims description 3
- 230000002159 abnormal effect Effects 0.000 description 14
- 238000011156 evaluation Methods 0.000 description 8
- 238000005338 heat storage Methods 0.000 description 8
- 238000012986 modification Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 239000006185 dispersion Substances 0.000 description 4
- 238000005485 electric heating Methods 0.000 description 3
- 238000001816 cooling Methods 0.000 description 2
- 238000013486 operation strategy Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of energy control, in particular to a comprehensive energy control method based on a neural network, which comprises the steps of drawing a time-energy storage curve graph according to screened effective energy storage information; drawing a pre-estimated energy storage curve; drawing an actual energy storage curve; integrating the area enclosed by the estimated energy storage curve and the actual energy storage curve in a coordinate system, and judging the estimated situation of the prediction unit for park energy storage based on the integral value; the control output unit outputs a pre-estimated qualified instruction or controls the output unit to convey a corresponding optimization instruction; and the control of comprehensive energy sources in the whole area is completed, and the energy storage information of each power generation device in the park is monitored so as to predict the subsequent energy storage change condition through a self-learning model, thereby improving the utilization rate of the energy sources in the park.
Description
Technical Field
The invention relates to the technical field of energy control, in particular to a comprehensive energy control method based on a neural network.
Background
Although the existing new energy source such as photovoltaic and wind power output prediction technologies are not few, the prediction precision is still lacking. At present, a centralized control method is generally adopted for energy management and scheduling control of a comprehensive energy system, and the energy efficiency and economy of equipment in the system are mathematically modeled, and the overall optimization target and an optimization iterative algorithm are combined for scheduling. However, in the face of multiple distributed energy sources, a large amount of control data and flexible and changeable control modes in the comprehensive energy source system, the centralized management is difficult to realize flexible and effective scheduling; therefore, the feasibility and practicality of the above method remain to be studied.
In order to cope with the increase of energy load and the large-scale application of various distributed energy sources, the comprehensive energy management and control system for the park brings the photovoltaic, fan, cold, hot and various adjustable loads in the park into the adjustment and control range, so that various energy sources and loads in the park need to be effectively monitored in real time, and the energy storage prediction, the energy consumption prediction condition and the like in the area are mastered, thereby improving the energy utilization rate of the whole park.
The existing park comprehensive energy management and control system is large in data acquisition quantity, large in fuzzy data and error data, large in energy management and control deviation and low in efficiency, and the actual conditions of energy storage prediction and energy consumption in the park cannot be accurately mastered.
Chinese patent publication No.: CN114322044B discloses a comprehensive energy system and its operation control method, including a control center, and an electric power supply subsystem, a heating subsystem and a cooling subsystem respectively connected with the control center; the heat supply subsystem comprises a medium-low temperature heat storage tank, a high-temperature heat storage tank and an electric heating boiler which are sequentially connected through pipelines; the high-temperature heat storage tank is used for performing heat exchange with the medium-low-temperature heat storage tank; the electric heating boiler is used for carrying out circulating heat exchange with the high-temperature heat storage tank; the control center is used for determining the operation strategy of the system according to the real-time load received in each period and a preset operation control model, and periodically controlling the power supply subsystem, the heat supply subsystem and the cooling subsystem according to the operation strategy. According to the system, the low-grade heat source recovered by the medium-low-temperature heat storage tank is stored in the high-temperature heat storage tank, and the electric heating boiler is utilized to circularly heat the high-temperature heat storage tank, so that the gradient temperature rise of the low-grade heat energy is realized; it follows that the prior art has the following problems: the energy storage in the park is not considered to be predicted, the data is not considered to be collected and screened, the data processing efficiency of energy management and control is affected, and the utilization rate of the energy in the park is further affected.
Disclosure of Invention
Therefore, the invention provides a comprehensive energy control method based on a neural network, which is used for solving the problems that in the prior art, prediction of energy storage in a park is not considered, data acquisition and screening are not considered, the data processing efficiency of energy management and control is affected, and the utilization rate of the energy in the park is further affected.
In order to achieve the above object, the present invention provides a comprehensive energy control method based on a neural network, including;
step S1, an acquisition unit acquires energy storage information of each power generation device in a park under a corresponding time node, a preprocessing unit screens the energy storage information output by the acquisition unit and draws a time-energy storage curve graph according to the screened effective energy storage information;
S2, inputting a time-energy storage curve graph drawn by the preprocessing unit into a self-learning model preset in the time-energy storage curve graph by the prediction unit so as to predict energy storage change conditions in a subsequent preset number of periods in the park based on the obtained energy storage conditions in the park, and drawing a predicted energy storage curve in the time-energy storage curve graph according to the predicted energy storage change conditions;
Step S3, when the prediction unit finishes drawing the estimated energy storage curve, the acquisition unit continuously acquires energy storage information of each power generation device in the operation process of the park, and the preprocessing unit performs screening processing on the acquired energy storage information and draws an actual energy storage curve in the time-energy storage curve graph based on the screened effective energy storage information;
S4, an analysis unit integrates the area enclosed by the estimated energy storage curve and the actual energy storage curve in a coordinate system and judges the estimated condition of the prediction unit for park energy storage based on the integral value, and when judging that the estimated condition of the prediction unit for park energy storage does not meet the standard, the analysis unit redetermines the estimated condition of the prediction unit for park energy storage based on the derivatives of the estimated energy storage curve and the actual energy storage curve or determines the reason that the estimated condition of the prediction unit for park energy storage does not meet the standard based on the integral value;
Step S5, the analysis unit controls the output unit to output a pre-estimated qualified instruction when judging that the pre-estimated energy storage of the prediction unit meets the standard, or controls the output unit to transmit a corresponding optimization instruction according to the determined reasons when judging that the pre-estimated energy storage of the prediction unit does not meet the standard, wherein the analysis unit comprises a screening optimization instruction for screening energy storage information of the preprocessing unit and a model optimization instruction for the self-learning model of the prediction unit;
and S6, the execution unit controls the corresponding unit to complete optimization according to the received instruction so as to complete the control of the comprehensive energy in the park.
Further, the analysis unit preliminarily judges that the estimated energy storage curve and the derivative curve of the actual energy storage curve are sequentially obtained when the estimated energy storage curve of the prediction unit for the park energy storage does not meet the standard based on the obtained integral value, and judges whether the estimated energy storage curve of the prediction unit for the park energy storage meets the standard or not based on the two derivative curves, or judges that the estimated energy storage curve of the prediction unit for the park energy does not meet the standard, and determines the reason of the non-meeting the standard based on the obtained integral value.
Further, the analysis unit marks the derivative curve of the estimated energy storage curve as an estimated derivative curve, marks the derivative curve of the actual energy storage curve as an actual derivative curve, and determines the estimated condition of the prediction unit for the park energy storage based on the coincidence degree of the estimated derivative curve and the actual derivative curve, and determines the treatment mode for the park when the estimated condition of the prediction unit for the park energy storage is determined to be not in accordance with the standard, or optimizes the screening standard of the preprocessing unit for the energy storage information;
and the coincidence ratio of the estimated derivative curve and the actual derivative curve is the ratio of the length of the coincidence line segment of the estimated derivative curve and the actual derivative curve to the total length of the estimated derivative curve.
Further, the processing manner of the analysis unit for the campus based on the contact ratio determination includes:
and optimizing a self-learning model in the analysis unit based on the electricity utilization conversion efficiency of each power generation device in the park, or determining an energy management and control scheme for the park based on the average energy storage condition of each electric device in the park.
Furthermore, the analysis unit is provided with a plurality of model correction modes aiming at the model output result of the self-learning model based on the average energy conversion rate of each power generation device, and the correction amplitudes of the model correction modes aiming at the model output result of the self-learning model are different.
Furthermore, the analysis unit is provided with a plurality of energy consumption correction modes aiming at the energy consumption standard of the energy consumption zone based on the average energy storage of all the electric equipment in the park, and the correction amplitudes of all the energy consumption correction modes aiming at the energy consumption standard are different.
Further, the analysis unit is provided with a plurality of discrete adjustment modes aiming at discrete value standards of the preprocessing unit screening data based on the overlap ratio difference value, and adjustment amplitudes of the discrete adjustment modes aiming at the discrete value standards are different;
and the preprocessing unit marks the energy storage value with the difference value larger than the discrete value standard as a discrete energy storage value and screens out the energy storage information corresponding to the discrete energy storage value.
Further, the analysis unit determines, based on the obtained integrated value, the reasons for the non-compliance with the criterion including the non-compliance of the energy storage of the campus and the non-compliance of the energy consumption zone;
The analysis unit determines a campus energy storage standard when it is determined that the campus energy storage is not in compliance with the standard, and determines that the energy consumption of the energy consumption zone is not in compliance with the energy consumption standard for the accurate energy consumption zone.
Further, the analysis unit is provided with a plurality of energy correction modes aiming at the energy consumption standard of the energy consumption area based on the difference value between the preset integral difference value and the integral difference value, and the correction amplitude of each energy correction mode aiming at the energy consumption standard is different.
Further, the analysis unit determines that the prediction unit predicts that the estimated energy storage of the park is not in accordance with the standard because the park energy storage is not in accordance with the standard, and determines the energy storage standard of the park according to the difference value between the preset integral difference value and the integral difference value.
Compared with the prior art, the method has the beneficial effects that the energy storage information of each power generation device in the park is monitored so as to predict the subsequent energy storage change condition through the self-learning model, thereby improving the utilization rate of energy sources in the park; and judging the prediction result of the self-learning model, so that when the prediction of the energy storage of the park is not in accordance with the standard, corresponding measures are taken, and the utilization rate of the energy of the park is further effectively improved while the accuracy of the prediction result of the energy storage is improved.
Further, the estimated energy storage curve is compared with the actual energy storage curve to judge whether the estimated energy storage of the park meets the standard according to the actual deviation of the two curves, namely, the integral value, when the deviation is smaller, the change trend of the estimated energy storage curve and the actual energy storage curve, namely, the estimated derivative curve and the actual derivative curve, is obtained respectively, when the change trend has abnormal deviation, the corresponding processing mode is determined, and the efficiency of energy management and control in the park is further effectively improved while the accuracy of the prediction result of the energy storage condition in the park is strictly controlled.
Further, the overlap ratio is further finely divided under the condition that the deviation of the change trend is smaller, specific processing measures are determined according to the division result, under the condition that the difference value of the first preset overlap ratio is too low, the change trend is only extremely small in deviation, the deviation between the estimated energy storage curve and the actual energy storage curve is also very small, the result output by the self-learning model is integrally adjusted at the moment, the deviation between the estimated energy storage curve and the actual energy storage curve is reduced, and the energy storage condition in a park is accurately predicted while the energy storage condition in the park is accurately controlled. When the overlap ratio gap is large, it is judged that energy storage has abnormal fluctuation due to the fact that energy consumption of the electricity utilization end is abnormal, so that an actual derivative curve has abnormal fluctuation, and therefore the energy consumption standard is adjusted to raise the electricity utilization limit standard of the electricity utilization end, and efficiency of energy management and control in a park is further effectively improved.
Further, under the condition that the deviation of the change trend is large, the actual deviation of the estimated energy storage curve and the actual energy storage curve is small, but the change trend of the two curves has large deviation fluctuation, so that the situation that the acquisition unit is abnormal for data acquisition is judged, the fluctuation of the prediction result of the self-learning model is caused by acquiring a large amount of data with large dispersion is judged, the acquisition standard of the data is improved aiming at the acquisition unit, namely, the discrete value standard is adjusted, continuous effective data with lower dispersion is selected for training, and the data is compared, so that the data processing amount of the preprocessing module is further reduced while a large amount of energy storage data is effectively screened.
Further, under the condition that the deviation between the estimated energy storage curve and the actual energy storage curve is large, the integral value is further divided, and under the condition that the difference between the integral value and the second preset integral value is large, the analysis unit is judged to be abnormal based on the evaluation standard of the integral value, and the obtained integral value and the evaluation interval deviate too much, so that the evaluation standard is adjusted, and the evaluation of data is more referential; under the condition that the difference value between the integral value and the second preset integral value is smaller, the integral value is still large at the moment, and the fact that the estimated energy storage curve deviates from the actual energy storage curve greatly due to the fact that the energy consumption of the power utilization end is greatly abnormal is judged, so that the energy consumption standard is adjusted, the energy storage condition in the park is accurately predicted, the energy storage and energy consumption conditions in the park are accurately controlled, a large amount of energy storage data are effectively screened, the data processing capacity of the preprocessing module is further reduced, the energy management and control deviation is effectively reduced, and meanwhile the energy management and control efficiency in the park is further effectively improved.
Drawings
FIG. 1 is a flow chart of steps of a method for controlling integrated energy based on a neural network according to an embodiment of the present invention;
FIG. 2 is a flowchart of an analysis unit determining whether a prediction unit predicts a park energy storage meeting a criterion based on an integrated value according to an embodiment of the present invention;
FIG. 3 is a flowchart of an energy consumption correction method for determining an energy consumption standard for an energy consumption area by an analysis unit according to an embodiment of the present invention based on an energy storage average value;
Fig. 4 is a flowchart of a discrete adjustment mode of the analysis unit for determining the discrete value standard of the screening data of the preprocessing unit based on the two-level overlap ratio difference value in the embodiment of the invention.
Detailed Description
In order that the objects and advantages of the invention will become more apparent, the invention will be further described with reference to the following examples; it should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
The above and further technical features and advantages of the present invention are described in more detail below with reference to the accompanying drawings.
Preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are merely for explaining the technical principles of the present invention, and are not intended to limit the scope of the present invention.
It should be noted that, in the description of the present invention, terms such as "upper," "lower," "left," "right," "inner," "outer," and the like indicate directions or positional relationships based on the directions or positional relationships shown in the drawings, which are merely for convenience of description, and do not indicate or imply that the apparatus or elements must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as limiting the present invention.
Furthermore, it should be noted that, in the description of the present invention, unless explicitly specified and limited otherwise, the terms "mounted," "connected," and "connected" are to be construed broadly, and may be either fixedly connected, detachably connected, or integrally connected, for example; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium, and can be communication between two elements. The specific meaning of the above terms in the present invention can be understood by those skilled in the art according to the specific circumstances.
Referring to fig. 1, fig. 2, fig. 3, and fig. 4, the method is a flowchart of steps of a comprehensive energy control method based on a neural network, a flowchart of an analysis unit determining whether a prediction unit predicts a park energy storage according to a standard based on an integral value, an energy consumption correction mode flowchart of an energy consumption standard for an energy consumption area based on an energy storage average value, and a discrete adjustment mode flowchart of a discrete value standard for screening data of a preprocessing unit based on a two-level overlap ratio difference value by an analysis unit according to an embodiment of the invention; the embodiment of the invention discloses a comprehensive energy control method based on a neural network, which comprises the following steps:
step S1, an acquisition unit acquires energy storage information of each power generation device in a park under a corresponding time node, a preprocessing unit screens the energy storage information output by the acquisition unit and draws a time-energy storage curve graph according to the screened effective energy storage information;
S2, inputting a time-energy storage curve graph drawn by the preprocessing unit into a self-learning model preset in the time-energy storage curve graph by the prediction unit so as to predict energy storage change conditions in a subsequent preset number of periods in the park based on the obtained energy storage conditions in the park, and drawing a predicted energy storage curve in the time-energy storage curve graph according to the predicted energy storage change conditions;
Step S3, when the prediction unit finishes drawing the estimated energy storage curve, the acquisition unit continuously acquires energy storage information of each power generation device in the operation process of the park, and the preprocessing unit performs screening processing on the acquired energy storage information and draws an actual energy storage curve in the time-energy storage curve graph based on the screened effective energy storage information;
S4, an analysis unit integrates the area enclosed by the estimated energy storage curve and the actual energy storage curve in a coordinate system and judges the estimated condition of the prediction unit for park energy storage based on the integral value, and when judging that the estimated condition of the prediction unit for park energy storage does not meet the standard, the analysis unit redetermines the estimated condition of the prediction unit for park energy storage based on the derivatives of the estimated energy storage curve and the actual energy storage curve or determines the reason that the estimated condition of the prediction unit for park energy storage does not meet the standard based on the integral value;
Step S5, the analysis unit controls the output unit to output a pre-estimated qualified instruction when judging that the pre-estimated energy storage of the prediction unit meets the standard, or controls the output unit to transmit a corresponding optimization instruction according to the determined reasons when judging that the pre-estimated energy storage of the prediction unit does not meet the standard, wherein the analysis unit comprises a screening optimization instruction for screening energy storage information of the preprocessing unit and a model optimization instruction for the self-learning model of the prediction unit;
and S6, the execution unit controls the corresponding unit to complete optimization according to the received instruction so as to complete the control of the comprehensive energy in the park.
Monitoring energy storage information of each power generation device in the park so as to predict the subsequent energy storage change condition through a self-learning model, thereby improving the utilization rate of energy sources in the park; and judging the prediction result of the self-learning model, so that when the prediction of the energy storage of the park is not in accordance with the standard, corresponding measures are taken, and the utilization rate of the energy of the park is further effectively improved while the accuracy of the prediction result of the energy storage is improved.
Specifically, the analysis unit determines whether or not the prediction unit's prediction of the campus energy storage meets a criterion based on the obtained integrated value:
If the integral value is smaller than or equal to a first preset integral value set in the analysis unit, the analysis unit judges that the prediction of the prediction unit for park energy storage accords with a standard, and controls the output unit to output a command for passing the prediction;
If the integral value is larger than the first preset integral value and smaller than or equal to a second preset integral value set in the analysis unit, the analysis unit preliminarily judges that the prediction of the prediction unit for the park energy storage does not accord with the standard, the analysis unit sequentially acquires derivative curves of the predicted energy storage curve and the actual energy storage curve, and judges whether the prediction of the prediction unit for the park energy storage accords with the standard again based on the two derivative curves;
And if the integral value is larger than the second preset integral value, the analysis unit judges that the prediction of the prediction unit for park energy storage does not meet the standard, and determines the reason of the non-meeting the standard based on the obtained integral value.
Specifically, when the analysis unit preliminarily judges that the prediction of the prediction unit for the park energy storage does not meet the standard, the analysis unit sequentially acquires the derivative curve of the predicted energy storage curve and the derivative curve of the actual energy storage curve, the analysis unit marks the derivative curve of the predicted energy storage curve as the predicted derivative curve, marks the derivative curve of the actual energy storage curve as the actual derivative curve, and redetermines the prediction condition of the prediction unit for the park energy storage based on the coincidence ratio of the predicted derivative curve and the actual derivative curve:
If the contact ratio is smaller than or equal to a first preset contact ratio set in the analysis unit, the analysis unit re-determines that the prediction of the prediction unit for energy storage in a park does not accord with a standard and optimizes the screening standard of the preprocessing unit for the energy storage information based on the difference value of the first preset contact ratio and the contact ratio;
If the contact ratio is larger than the first preset contact ratio and smaller than or equal to the second preset contact ratio set in the analysis unit, the analysis unit re-determines that the prediction of the prediction unit for the park energy storage does not accord with the standard, and determines a park processing mode based on the difference value of the contact ratio and the first preset contact ratio;
if the overlap ratio is larger than the second preset overlap ratio, the analysis unit re-determines that the prediction of the prediction unit for park energy storage accords with a standard;
And the coincidence ratio of the estimated derivative curve and the actual derivative curve is the ratio of the length of the coincidence line segment of the estimated derivative curve and the actual derivative curve to the total length of the estimated derivative curve.
The estimated energy storage curve is compared with the actual energy storage curve to judge whether the estimated energy storage of the park meets the standard according to the actual deviation of the two curves, namely, the integral value, when the deviation is smaller, the change trend of the estimated energy storage curve and the actual energy storage curve, namely, the estimated derivative curve and the actual derivative curve, is respectively obtained, when the change trend has abnormal deviation, the corresponding processing mode is determined, and the efficiency of energy management and control in the park is further effectively improved while the accuracy of the estimated result of the energy storage condition in the park is strictly controlled.
Specifically, the analysis unit marks the difference between the contact ratio and the first preset contact ratio as a first-level contact ratio difference value and determines an energy management and control scheme for the park according to the first-level contact ratio difference value:
If the first-order overlap ratio difference value is smaller than or equal to the preset first-order overlap ratio difference value set in the analysis unit, the analysis unit optimizes a self-learning model in the analysis unit based on the distribution condition of the power utilization conversion efficiency of each power generation device in the park;
and if the first-level overlap ratio difference value is larger than the preset first-level overlap ratio difference value, the analysis unit determines an energy management and control scheme for the park based on the average energy storage condition of each electric equipment in the park.
The method comprises the steps of carrying out further fine division on the contact ratio under the condition that the deviation of the change trend is smaller, determining specific processing measures according to division results, under the condition that the difference value of the first preset contact ratio is too low, only minimal deviation exists on the change trend, the deviation between an estimated energy storage curve and an actual energy storage curve is small, carrying out integral adjustment on the output result of a self-learning model at the moment, so as to reduce the deviation between the estimated energy storage curve and the actual energy storage curve, and accurately predicting the energy storage condition in a park and simultaneously accurately controlling the energy storage and energy consumption condition in the park. When the overlap ratio gap is large, it is judged that energy storage has abnormal fluctuation due to the fact that energy consumption of the electricity utilization end is abnormal, so that an actual derivative curve has abnormal fluctuation, and therefore the energy consumption standard is adjusted to raise the electricity utilization limit standard of the electricity utilization end, and efficiency of energy management and control in a park is further effectively improved.
Specifically, the analysis unit obtains the energy conversion rate of each power generation device when the first-order overlap ratio difference value is smaller than or equal to the preset first-order overlap ratio difference value, and determines a model correction mode for the output result of the self-learning model based on an average value of the energy conversion rates, wherein:
The first model correction mode is that the analysis unit corrects the result output by the self-learning model by using a first preset model correction coefficient; the first model correction mode meets the condition that the average energy conversion rate is smaller than or equal to a first preset average energy conversion rate set in the analysis unit;
The second model correction mode is that the analysis unit corrects the result output by the self-learning model by using a second preset model correction coefficient; the second model correction mode meets the condition that the average energy conversion rate is smaller than or equal to a second preset average energy conversion rate set in the analysis unit and larger than the first preset average energy conversion rate, and the first preset average energy conversion rate is smaller than the second preset average energy conversion rate;
The third model correction mode is that the analysis unit corrects the result output by the self-learning model by using a third preset model correction coefficient; the third model modification mode satisfies that the average energy conversion rate is greater than the second preset average energy conversion rate.
Specifically, the analysis unit calculates an energy storage average value of each electric device when the first-order overlap ratio difference value is larger than the preset first-order overlap ratio difference value, and determines an energy consumption correction mode of an energy consumption standard for an energy consumption area based on the obtained energy storage average value, wherein:
The first energy consumption correction mode is that the analysis unit corrects the energy consumption standard to a corresponding value by using a first preset energy consumption correction coefficient; the first energy consumption correction mode meets the condition that the energy storage average value is smaller than or equal to a first preset energy storage average value set in the analysis unit;
The second energy consumption correction mode is that the analysis unit corrects the energy consumption standard to a corresponding value by using a second preset energy consumption correction coefficient; the second energy consumption correction mode meets the condition that the energy storage average value is smaller than or equal to a second preset energy storage average value set in the analysis unit and larger than the first preset energy storage average value, and the first preset energy storage average value is smaller than the second preset energy storage average value;
the third energy consumption correction mode is that the analysis unit corrects the energy consumption standard to a corresponding value by using a third preset energy consumption correction coefficient; the third energy consumption correction mode meets the condition that the energy storage average value is larger than the second preset energy storage average value.
Specifically, the analysis unit marks the difference between the first preset contact ratio and the contact ratio as a second-level contact ratio difference, and determines a discrete adjustment mode of a discrete value standard of screening data for the preprocessing unit according to the second-level contact ratio difference, wherein:
the first discrete adjustment mode is that the analysis unit uses a first preset discrete adjustment coefficient to adjust the discrete value standard to a corresponding value; the first discrete adjustment mode meets the condition that the second-level overlap ratio difference value is smaller than or equal to a first preset second-level overlap ratio set in the analysis unit;
The second discrete adjustment mode is that the analysis unit uses a second preset discrete adjustment coefficient to adjust the discrete value standard to a corresponding value; the second discrete adjustment mode meets the condition that the second-level contact ratio difference value is smaller than or equal to a second preset second-level contact ratio set in the analysis unit and larger than the first preset second-level contact ratio, and the first preset second-level contact ratio is smaller than the second preset second-level contact ratio;
the third discrete adjustment mode is that the analysis unit uses a third preset discrete adjustment coefficient to adjust the discrete value standard to a corresponding value; the third discrete adjustment mode satisfies that the second-level overlap ratio difference is greater than the second preset second-level overlap ratio.
Under the condition that the deviation of the change trend is large, the actual deviation of the estimated energy storage curve and the actual energy storage curve is small, but the change trend of the two curves has large deviation fluctuation, so that the situation that the collection unit is abnormal for data collection and the collection of a large amount of data with large dispersion causes fluctuation of the prediction result of the self-learning model is judged, the collection standard of the data is improved for the collection unit, namely, the discrete value standard is adjusted, continuous effective data with lower dispersion is selected for training, and the data is compared, and the data processing capacity of the preprocessing module is further reduced while a large amount of energy storage data is effectively screened.
Specifically, the analysis unit calculates a difference value between the integrated value and a second preset integrated value when the integrated value is greater than the second preset integrated value, and records the obtained difference value as an integrated difference value, and the analysis unit determines a reason judgment mode that the prediction unit does not accord with a standard for the prediction of the park energy storage based on the integrated difference value, wherein:
the first reason judging mode is that the analysis unit judges that the prediction unit judges that the estimated energy storage of the park does not meet the standard because the park energy storage does not meet the standard, and the park energy storage standard is determined according to the difference value between the preset integral difference value and the integral difference value; the first cause judgment mode satisfies that the integral difference value is larger than a preset integral difference value set in the analysis unit;
The second reason judging mode is that the analysis unit judges that the prediction unit judges that the estimated energy storage of the park is not in accordance with the standard because the energy consumption of the energy consumption area is not in accordance with the standard, and the energy consumption standard of the energy consumption area is determined according to the difference value of the integral difference value and the preset integral difference value; the second cause judgment mode satisfies that the integral difference value is smaller than or equal to the preset integral difference value.
Dividing the integral value further under the condition that the deviation between the estimated energy storage curve and the actual energy storage curve is large, and judging that the obtained integral value deviates too much from an evaluation interval due to the fact that the analysis unit is abnormal based on the evaluation standard of the integral value under the condition that the difference between the integral value and a second preset integral value is large, so that the evaluation standard is adjusted to enable the evaluation of data to be more referential; under the condition that the difference value between the integral value and the second preset integral value is smaller, the integral value is still large at the moment, and the fact that the estimated energy storage curve deviates from the actual energy storage curve greatly due to the fact that the energy consumption of the power utilization end is greatly abnormal is judged, so that the energy consumption standard is adjusted, the energy storage condition in the park is accurately predicted, the energy storage and energy consumption conditions in the park are accurately controlled, a large amount of energy storage data are effectively screened, the data processing capacity of the preprocessing module is further reduced, the energy management and control deviation is effectively reduced, and meanwhile the energy management and control efficiency in the park is further effectively improved.
Specifically, the analysis unit marks the difference between the preset integral difference and the integral difference as a low secondary difference and determines an energy correction mode of an energy consumption standard for the energy consumption area according to the low secondary difference, wherein:
The first energy correction mode is that the analysis unit corrects the energy consumption standard to a corresponding value by using a first preset energy correction coefficient; the first energy source correction mode meets the condition that the low secondary difference value is smaller than or equal to a first preset low secondary difference value set in the analysis unit;
The second energy correction mode is that the analysis unit corrects the energy consumption standard to a corresponding value by using a second preset energy correction coefficient; the second energy source correction mode meets the condition that the low secondary difference value is smaller than or equal to a second preset low secondary difference value set in the analysis unit and larger than the first preset low secondary difference value, and the first preset low secondary difference value is smaller than the second preset low secondary difference value;
the third energy correction mode is that the analysis unit corrects the energy consumption standard to a corresponding value by using a third preset energy correction coefficient; the third energy correction mode satisfies that the low secondary difference value is larger than a second preset low secondary difference value set in the analysis unit.
Specifically, the analysis unit records the difference between the integral difference and the preset integral difference as a high secondary difference, and determines a standard adjustment mode for the first preset integral value and the second preset integral value according to the low secondary difference, wherein:
The first standard adjustment mode is that the analysis unit adjusts the first preset integral value and the second preset integral value to corresponding values by using a first preset standard adjustment coefficient; the first standard adjustment mode meets the condition that the high secondary difference value is smaller than or equal to a first preset high secondary difference value set in the analysis unit;
the second standard adjustment mode is that the analysis unit adjusts the first preset integral value and the second preset integral value to corresponding values by using a second preset standard adjustment coefficient; the second standard adjustment mode meets the condition that the high-level difference value is smaller than or equal to a second preset high-level difference value set in the analysis unit and larger than the first preset high-level difference value, and the first preset high-level difference value is smaller than the second preset high-level difference value;
The third standard adjustment mode is that the analysis unit adjusts the first preset integral value and the second preset integral value to corresponding values by using a third preset standard adjustment coefficient; the third standard adjustment mode satisfies that the high secondary difference value is larger than a second preset high secondary difference value set in the analysis unit.
Thus far, the technical solution of the present invention has been described in connection with the preferred embodiments shown in the drawings, but it is easily understood by those skilled in the art that the scope of protection of the present invention is not limited to these specific embodiments. Equivalent modifications and substitutions for related technical features may be made by those skilled in the art without departing from the principles of the present invention, and such modifications and substitutions will be within the scope of the present invention.
The foregoing description is only of the preferred embodiments of the invention and is not intended to limit the invention; various modifications and variations of the present invention will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. The comprehensive energy control method based on the neural network is characterized by comprising the following steps of:
step S1, an acquisition unit acquires energy storage information of each power generation device in a park under a corresponding time node, a preprocessing unit screens the energy storage information output by the acquisition unit and draws a time-energy storage curve graph according to the screened effective energy storage information;
S2, inputting a time-energy storage curve graph drawn by the preprocessing unit into a self-learning model preset in the time-energy storage curve graph by the prediction unit so as to predict energy storage change conditions in a subsequent preset number of periods in the park based on the obtained energy storage conditions in the park, and drawing a predicted energy storage curve in the time-energy storage curve graph according to the predicted energy storage change conditions;
Step S3, when the prediction unit finishes drawing the estimated energy storage curve, the acquisition unit continuously acquires energy storage information of each power generation device in the operation process of the park, and the preprocessing unit performs screening processing on the acquired energy storage information and draws an actual energy storage curve in the time-energy storage curve graph based on the screened effective energy storage information;
S4, an analysis unit integrates the area enclosed by the estimated energy storage curve and the actual energy storage curve in a coordinate system and judges the estimated condition of the prediction unit for park energy storage based on the integral value, and when judging that the estimated condition of the prediction unit for park energy storage does not meet the standard, the analysis unit redetermines the estimated condition of the prediction unit for park energy storage based on the derivatives of the estimated energy storage curve and the actual energy storage curve or determines the reason that the estimated condition of the prediction unit for park energy storage does not meet the standard based on the integral value;
Step S5, the analysis unit controls the output unit to output a pre-estimated qualified instruction when judging that the pre-estimated energy storage of the prediction unit meets the standard, or controls the output unit to transmit a corresponding optimization instruction according to the determined reasons when judging that the pre-estimated energy storage of the prediction unit does not meet the standard, wherein the analysis unit comprises a screening optimization instruction for screening energy storage information of the preprocessing unit and a model optimization instruction for the self-learning model of the prediction unit;
and S6, the execution unit controls the corresponding unit to complete optimization according to the received instruction so as to complete the control of the comprehensive energy in the park.
2. The method for controlling integrated energy based on a neural network according to claim 1, wherein the analysis unit initially determines, based on the obtained integrated value, whether the prediction unit's prediction for park energy storage does not meet a criterion, sequentially obtaining the predicted energy storage curve and the derivative curve of the actual energy storage curve, and re-determining, based on the two derivative curves, whether the prediction unit's prediction for park energy storage meets the criterion,
Or when the prediction unit is judged to be inconsistent with the standard for the prediction of the park energy storage, determining the reason of the inconsistent with the standard based on the obtained integral value.
3. The neural network-based comprehensive energy control method according to claim 2, wherein the analysis unit marks a derivative curve of the estimated energy storage curve as an estimated derivative curve and marks a derivative curve of the actual energy storage curve as an actual derivative curve, the analysis unit redetermines the estimated condition of the prediction unit for the park energy storage based on the coincidence of the estimated derivative curve and the actual derivative curve, and determines a park-oriented processing mode when the redetermined estimate of the prediction unit for the park energy storage does not meet a standard, or optimizes a screening standard of the preprocessing unit for energy storage information;
and the coincidence ratio of the estimated derivative curve and the actual derivative curve is the ratio of the length of the coincidence line segment of the estimated derivative curve and the actual derivative curve to the total length of the estimated derivative curve.
4. The neural network-based integrated energy control method of claim 3, wherein the analysis unit determines the manner of processing for the campus based on the overlap ratio comprises:
and optimizing a self-learning model in the analysis unit based on the electricity utilization conversion efficiency of each power generation device in the park, or determining an energy management and control scheme for the park based on the average energy storage condition of each electric device in the park.
5. The neural network-based comprehensive energy control method according to claim 4, wherein the analysis unit is provided with a plurality of model correction modes for the model output result of the self-learning model based on the average energy conversion rate of each power generation device, and the correction amplitudes of the model correction modes for the model output result of the self-learning model are different.
6. The method for controlling integrated energy based on a neural network according to claim 4, wherein the analysis unit is provided with a plurality of energy consumption correction modes aiming at energy consumption standards of the energy consumption area based on average energy storage of electric equipment in the campus, and correction amplitudes of the energy consumption correction modes aiming at the energy consumption standards are different.
7. The neural network-based comprehensive energy control method according to claim 3, wherein the analysis unit is provided with a plurality of discrete adjustment modes aiming at discrete value standards of the preprocessing unit screening data based on the coincidence degree difference value, and adjustment amplitudes of the discrete adjustment modes aiming at the discrete value standards are different;
and the preprocessing unit marks the energy storage value with the difference value larger than the discrete value standard as a discrete energy storage value and screens out the energy storage information corresponding to the discrete energy storage value.
8. The neural network-based integrated energy control method of claim 6, wherein the analysis unit determining a cause of the non-compliance with the criterion based on the obtained integrated value includes the non-compliance of the campus energy storage with the criterion and the non-compliance of the energy consumption region;
The analysis unit determines a campus energy storage standard when it is determined that the campus energy storage is not in compliance with the standard, and determines that the energy consumption of the energy consumption zone is not in compliance with the energy consumption standard for the accurate energy consumption zone.
9. The neural network-based comprehensive energy control method according to claim 8, wherein the analysis unit is provided with a plurality of energy correction modes aiming at the energy consumption standard of the energy consumption area based on a preset integral difference value and a difference value of the integral difference value, and correction amplitudes of the energy correction modes aiming at the energy consumption standard are different.
10. The neural network-based integrated energy control method of claim 8, wherein the analysis unit determines that the prediction unit is not in compliance with the criteria for the prediction of the campus energy storage is because the campus energy storage is not in compliance with the criteria, and determines the criteria for the campus energy storage according to a difference between a preset integral difference and an integral difference.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410129667.5A CN117674301B (en) | 2024-01-31 | 2024-01-31 | Comprehensive energy control method based on neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410129667.5A CN117674301B (en) | 2024-01-31 | 2024-01-31 | Comprehensive energy control method based on neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117674301A CN117674301A (en) | 2024-03-08 |
| CN117674301B true CN117674301B (en) | 2024-05-17 |
Family
ID=90079229
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410129667.5A Active CN117674301B (en) | 2024-01-31 | 2024-01-31 | Comprehensive energy control method based on neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117674301B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113255992A (en) * | 2021-05-21 | 2021-08-13 | 南京南瑞继保电气有限公司 | Optimal scheduling method and device for park-level multi-energy complementary system |
| CN114218856A (en) * | 2021-12-09 | 2022-03-22 | 四川艾德瑞电气有限公司 | Method for judging abnormity of contact network based on autoregression and deep learning model |
| CN116454928A (en) * | 2023-02-21 | 2023-07-18 | 国网浙江省电力有限公司嘉兴供电公司 | Multi-type energy storage cooperative scheduling method considering multiple time scales |
| KR20230161721A (en) * | 2022-05-19 | 2023-11-28 | 세종대학교산학협력단 | Method for predicting renewable energy consumption and production based on deep learning and apparatus thereof |
-
2024
- 2024-01-31 CN CN202410129667.5A patent/CN117674301B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113255992A (en) * | 2021-05-21 | 2021-08-13 | 南京南瑞继保电气有限公司 | Optimal scheduling method and device for park-level multi-energy complementary system |
| CN114218856A (en) * | 2021-12-09 | 2022-03-22 | 四川艾德瑞电气有限公司 | Method for judging abnormity of contact network based on autoregression and deep learning model |
| KR20230161721A (en) * | 2022-05-19 | 2023-11-28 | 세종대학교산학협력단 | Method for predicting renewable energy consumption and production based on deep learning and apparatus thereof |
| CN116454928A (en) * | 2023-02-21 | 2023-07-18 | 国网浙江省电力有限公司嘉兴供电公司 | Multi-type energy storage cooperative scheduling method considering multiple time scales |
Non-Patent Citations (1)
| Title |
|---|
| 基于模型预测控制的储能平抑光伏波动的控制策略;张德隆 等;电器与能效管理技术;20160730(2016年第14期);第34-40页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117674301A (en) | 2024-03-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Oshnoei et al. | Robust control scheme for distributed battery energy storage systems in load frequency control | |
| Gao et al. | Technologies in smart district heating system | |
| CN108490790A (en) | A kind of overheating steam temperature active disturbance rejection cascade control method based on multiple-objection optimization | |
| WO2011124720A2 (en) | Clouds managing system for a solar field, method for operating the clouds management system and solar field with the clouds managing system | |
| US20110245986A1 (en) | System and method for multi-objective management of the electrical and thermal energy generated by a co/trigeneration energy system in a multi-source energy plant | |
| CN111102645B (en) | Heat balance control method for central heating system | |
| CN107785929B (en) | Solar thermal power station power prediction system and operation method | |
| CN108258684A (en) | A kind of clean energy resource power grid " source lotus domain " coordinates regulation and control method | |
| WO2023083235A1 (en) | Optimization method and system for performing prediction on power distribution network in distributed manner | |
| CN113110057A (en) | Heating power station energy-saving control method based on artificial intelligence and intelligent decision system | |
| CN116388278A (en) | Micro-grid group cooperative control method, device, equipment and medium | |
| CN113110056B (en) | Heat supply intelligent decision-making method and intelligent decision-making machine based on artificial intelligence | |
| CN117674301B (en) | Comprehensive energy control method based on neural network | |
| CN117273284A (en) | Abnormal data monitoring system for enterprise electricity balance | |
| CN116499023A (en) | Intelligent control method and system for geothermal coupling solar heating station | |
| CN115952695A (en) | Temperature detection data remote interaction method based on heat supply control system | |
| CN111664502B (en) | Correction method of room temperature correction system | |
| CN112467805A (en) | Double-asynchronous feedback thermal power generation primary and secondary frequency modulation comprehensive optimization system | |
| Ma et al. | Controller design and parameter optimization of aggregated thermostatically controlled loads for frequency regulation | |
| CN115224741B (en) | Distributed photovoltaic low-voltage grid-connected adjusting system and method | |
| CN117254532B (en) | Cluster regulation and control method and system for distributed photovoltaic access power distribution network | |
| CN219955479U (en) | Intelligent optimizing system of heat supply circulating pump | |
| CN117558944A (en) | Proton exchange membrane fuel cell temperature regulating system and method | |
| CN117595366A (en) | Energy regulation and control system is synthesized in garden | |
| CN117559426A (en) | Multi-type power supply and new energy integrated control method based on self-organizing map network |
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 |