CN115912352A - Source-load power prediction method suitable for photovoltaic ship - Google Patents
Source-load power prediction method suitable for photovoltaic ship Download PDFInfo
- Publication number
- CN115912352A CN115912352A CN202211531908.6A CN202211531908A CN115912352A CN 115912352 A CN115912352 A CN 115912352A CN 202211531908 A CN202211531908 A CN 202211531908A CN 115912352 A CN115912352 A CN 115912352A
- Authority
- CN
- China
- Prior art keywords
- power
- ship
- photovoltaic
- load
- data
- 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.)
- Pending
Links
Images
Landscapes
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention relates to the technical field of ship electric power, and particularly discloses a 'source-load' power prediction method suitable for a photovoltaic ship, which comprises the following steps: s1, constructing a source-load power prediction model of a hybrid power system by analyzing power fluctuation rules of the power generation power of a marine photovoltaic power generation system and the power fluctuation rules of ship loads; s2, establishing a prediction model according to the step S1, and constructing a hybrid power system electromagnetic transient simulation model, wherein the hybrid power system electromagnetic transient simulation model comprises photovoltaic power fluctuation, ship load and the like; according to the method, a source-load power prediction model of a hybrid power system is researched and constructed by deeply analyzing power fluctuation rules of a marine photovoltaic power generation system and ship loads; and then, constructing an electromagnetic transient simulation model of the hybrid power system, and revealing the influence mechanism of photovoltaic power generation power fluctuation, ship load high-power sudden change, diesel generator set start and stop, three-phase voltage normal unbalance and short-circuit fault on the hybrid power system.
Description
Technical Field
The invention relates to the technical field of ship electric power, in particular to a source-load power prediction method suitable for a photovoltaic ship.
Background
In the energy structure of the world, primary energy utilized by human beings is mainly fossil energy, for example, petroleum, ship and ship power stations mainly generate electricity by burning heavy oil, so that the power generation cost is high, and in addition, in annual carbon and nitrogen emissions in the world, a certain proportion is caused by the emission of the marine shipping industry. Sustainable energy such as solar energy and the like and traditional ship power generation equipment are mixed to supply power to form a new trend of a ship power supply system, the comparison of the advantages and the disadvantages of the two power stations in the aspect of economy has practical significance, the economy evaluation is carried out on a ship hybrid power system provided with solar photovoltaic equipment, and scientific and reasonable analysis needs to be carried out on the applicability and the durability on the basis of comprehensive calculation and evaluation of solar radiation intensity, the investment cost of the photovoltaic system and the fuel price so as to reasonably avoid potential investment risks. The longitude and latitude of the ship in the process of sailing can change constantly, and meanwhile, the ship has a swinging phenomenon, so that the actual receiving quantity of the irradiance of the photovoltaic panel in the ship is different from that of the photovoltaic panel in the traditional land, and the output power of the photovoltaic system of the ship is different from that of the land system. At present, renewable energy is used as all or part of power supply equipment of a ship, most of the renewable energy is used for reference and even application of experience of a land photovoltaic system, and the particularity of a photovoltaic power generation device on the ship is not considered, namely the influence of comprehensive motion of the ship on the photovoltaic power generation equipment is not considered, wherein six degrees of freedom influencing the motion of the ship are transverse oscillation, longitudinal oscillation, sinking and floating, transverse oscillation, longitudinal oscillation and yawing.
In the green ship with the photovoltaic, the photovoltaic power generation system lacks inertia, the ship power system is also an autonomous power system with relatively small capacity, the total load of the ship is close to the total power capacity, and when the working condition of the ship is complex and changeable, the power fluctuation is frequent. For the ship power system integrated with DG, the brittleness of the ship power system is increased due to the random fluctuation of new energy, and the reliability of the system is reduced. On the one hand, on the new energy ships, the fluctuation of output electric energy is stabilized mainly by additionally arranging an energy storage device or arranging a selectable electric load. Therefore, the impact of new energy on a ship power system is reduced, the safety and the stability of the operation of the power system are guaranteed, and a certain effect is achieved. However, there are certain problems: if the fluctuations are all stabilized by energy storage, the fluctuations can possibly fall into frequent charging and discharging, and the service life of the energy storage is greatly influenced, so that the economic cost of the solar ship is increased; on the other hand, the photovoltaic power generation system outputs electric energy through the power electronic conversion device, the traditional control strategies comprise droop control, constant-frequency constant-voltage control and PQ control, and the droop control, the constant-frequency constant-voltage control and the PQ control all depend on a phase-locked loop to track the frequency of a power grid, so that the balance of the output power of the system and the photovoltaic array is kept in real time, and the response is quick. However, the photovoltaic power generation system is lack of sufficient inertia, and the disturbance resistance is reduced. The energy system power cannot be accurately predicted.
The invention discloses a multi-energy ship control management method and a device based on a load prediction algorithm in the prior art (CN 202111493950.9), and the method comprises the following steps: acquiring load data and temperature, humidity and climate data, and preprocessing the acquired data to determine a training sample set; establishing a load prediction model by adopting a least square support vector machine algorithm; training a load prediction model based on a training sample set, performing parameter optimization through a particle swarm algorithm, and determining an optimal load prediction model; acquiring current load data, on-off state data and temperature, humidity and climate data, and inputting the current load data and the temperature, humidity and climate data into an optimal load prediction model to obtain a ship load prediction value; acquiring ship power parameters to obtain input power; and controlling the electric propulsion of the ship, the energy equipment of the ship and the switching states of the first-stage load, the second-stage load and the third-stage load based on the input power, the predicted ship load value and the switching state data. The invention can grasp the change condition of the ship load in time, increase or close the diesel generator, and reduce the operation cost. However, the method is only limited to the prediction of the ship load, has defects in the actual use process of the photovoltaic power generation ship, and cannot judge the influence mechanism of the photovoltaic power generation power fluctuation, the high-power sudden change of the ship load, the start and stop of the diesel generator set, the normal imbalance of the three-phase voltage and the short-circuit fault on the hybrid power system.
In order to further optimize the power control of the ship, and particularly apply to a system with small relative capacity, such as a ship, the invention provides a 'source-load' power prediction method suitable for a photovoltaic ship.
Disclosure of Invention
The invention aims to provide a source-load power prediction method suitable for a photovoltaic ship, wherein a CNN-LSTM mixed deep neural network is adopted to construct a marine photovoltaic power generation power prediction model so as to realize the prediction of the marine photovoltaic power generation power; and the power distribution of the ship load in the three-phase circuit, the starting frequency and the running time length of the ship load are determined by combining with the 'power load calculation book' of the ship, and the load requirements of different running working conditions are combined to analyze the step dynamic fluctuation characteristics of the ship load and the unbalance degree of the ship load in the three-phase circuit. Researching and constructing a ship power load three-phase power prediction model number based on a ship navigation plan; the method comprises the following steps of researching and constructing a source-load power prediction model of the hybrid power system by deeply analyzing power fluctuation rules of a photovoltaic power generation system for a ship and ship loads; then, an electromagnetic transient simulation model of the hybrid power system is constructed, and the influence mechanism of photovoltaic power generation power fluctuation, ship load high-power sudden change, diesel generating set start and stop, three-phase voltage normal unbalance and short-circuit fault on the hybrid power system can be revealed.
In order to achieve the purpose, the invention provides the following technical scheme: a method of 'source-load' power prediction for a photovoltaic vessel, the method comprising the steps of:
s1, constructing a source-load power prediction model of a hybrid power system by analyzing power fluctuation rules of the power generation power of a marine photovoltaic power generation system and the power fluctuation rules of ship loads;
step S2, establishing a prediction model according to the step S1, and constructing a hybrid electric power system electromagnetic transient simulation model, wherein the hybrid electric power system electromagnetic transient simulation model comprises data information of photovoltaic power generation power fluctuation, ship load high-power sudden change, diesel generating set starting and stopping, three-phase voltage normal unbalance and short-circuit fault, and is used for predicting the influence mechanism of the photovoltaic power generation power fluctuation, the ship load high-power sudden change, the diesel generating set starting and stopping, the three-phase voltage normal unbalance and the short-circuit fault on the hybrid electric power system;
and S3, combining a source-load power prediction model of the hybrid power system and an electromagnetic transient simulation model of the hybrid power system, and outputting a prediction result.
As a preferred embodiment of the invention, in the method, the generation power prediction of the photovoltaic power generation system is to combine the photoelectric conversion nonlinear characteristic of the solar cell, consider the influence of ship speed, course, geographical position, ship swinging angle, cloud cover and upper deck building shielding, deck surface temperature radiation on the photovoltaic power generation, analyze the spatial correlation among factors and the correlation of single-factor time sequence data, and adopt a CNN-LSTM mixed deep neural network to construct a marine photovoltaic power generation power prediction model.
The method for forecasting the ship load power is a preferred embodiment of the invention, wherein the forecasting of the ship load power in the method is to determine the power distribution of the ship load in a three-phase circuit, the starting frequency and the running time length of the ship load according to a power load calculation book of a ship, analyze the step dynamic fluctuation characteristics of the ship load and the unbalance degree of the ship load in the three-phase circuit by combining the load demands of different running conditions, and construct a ship power load three-phase power forecasting model based on a ship navigation plan.
As a preferred embodiment of the present invention, in the step S1, the building step of the photovoltaic power generation prediction model for the ship comprises:
firstly, the CNN model is used for extracting the spatial characteristics of the photovoltaic data, then the LSTM model is used for extracting the temporal characteristics of the photovoltaic data on the basis of the extracted spatial characteristics, and then the photovoltaic power is predicted.
As a preferred embodiment of the invention, the marine photovoltaic power generation power prediction model construction method in step S1 is as follows:
firstly, collecting potential influence factor data which can influence photovoltaic power generation power by a real ship, then preprocessing the data, modifying and filling the data, and then adopting a normalization formula
Normalizing the data;
dividing the processed data into a training set, a verification set and a test set, wherein the training set, the verification set and the test set are not crossed;
photovoltaic data is then input into the vector
Inputting the data to an upper CNN model to obtain the spatial characteristics of the photovoltaic data; the lower LSTM model stores time information about important photovoltaic data features extracted through CNN;
and finally, obtaining the optimal lag time step, the number of layers of each model in the hybrid model and the parameters of each layer through a large amount of data training, and realizing the construction of the photovoltaic power generation power prediction model for the CNN-LSTM ship.
In a preferred embodiment of the present invention, the influencing factors include external environmental factors and navigation attitude data.
In a preferred embodiment of the present invention, the data patch is processed by linear interpolation or weighted moving average.
Compared with the prior art, the invention has the beneficial effects that:
through deep analysis of power fluctuation laws of a marine photovoltaic power generation system and a marine load, a source-load power prediction model of a hybrid power system is researched and constructed, an electromagnetic transient simulation model of the hybrid power system is constructed, and the influence mechanism of photovoltaic power generation power fluctuation, marine load high-power sudden change, diesel generator set starting and stopping, three-phase voltage normal unbalance and short-circuit fault on the hybrid power system is revealed. Through the exploration, the energy utilization and power system optimization design of the photovoltaic ship during sailing can be greatly referred to.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention.
FIG. 1 is a flow chart of a method of the present invention for "source-load" power prediction for a photovoltaic vessel;
FIG. 2 is a CNN-LSTM model structure diagram of the 'source-load' power prediction method applicable to photovoltaic ships in accordance with the present invention;
FIG. 3 is a flow chart of a CNN-LSTM marine photovoltaic power generation power prediction model construction of a 'source-load' power prediction method applicable to a photovoltaic ship according to the present invention;
FIG. 4 is a flow chart of a ship load power prediction model construction of the source-load power prediction method applicable to photovoltaic ships.
Detailed Description
In order to make the technical problems, technical solutions and advantageous effects to be solved by the present invention more clearly apparent, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Referring to fig. 1-4, in order to achieve the above object, the present invention provides the following technical solutions: a method of 'source-load' power prediction for a photovoltaic vessel, the method comprising the steps of:
s1, constructing a source-load power prediction model of a hybrid power system by analyzing power fluctuation rules of the power generation power of a marine photovoltaic power generation system and the power fluctuation rules of ship loads;
step S2, establishing a prediction model according to the step S1, and establishing a hybrid electric power system electromagnetic transient simulation model, wherein the hybrid electric power system electromagnetic transient simulation model comprises data information of photovoltaic power generation power fluctuation, ship load high-power sudden change, diesel generating set start-stop, three-phase voltage normal unbalance and short-circuit fault, and is used for predicting the influence mechanism of the photovoltaic power generation power fluctuation, the ship load high-power sudden change, the diesel generating set start-stop, the three-phase voltage normal unbalance and the short-circuit fault on the hybrid electric power system;
and S3, combining a source-load power prediction model of the hybrid power system and an electromagnetic transient simulation model of the hybrid power system, and outputting a prediction result.
Furthermore, in the method, the generation power prediction of the photovoltaic power generation system is to combine the photoelectric conversion nonlinear characteristic of the solar cell, consider the influence of ship speed, course, geographical position, ship swing angle, cloud layer and upper deck building shielding, deck surface temperature radiation on the photovoltaic power generation power, analyze the spatial correlation among factors and the correlation of single-factor time sequence data, and construct a marine photovoltaic power generation power prediction model by adopting a CNN-LSTM mixed deep neural network.
The CNN + LSTM is a space-time network, wherein a Convolutional Neural Network (CNN) is a special artificial Neural network, which is different from other models of Neural Networks (e.g., BP Neural network, RNN Neural network, etc.), and the most important feature of the Convolutional Neural network is Convolutional operations (Convolutional operations). CNN is therefore excellent for many applications, in particular image-related tasks, such as computer vision problems like image classification, image semantic segmentation, image retrieval, object detection, etc.
Convolutional neural networks are suitable for processing spatial data, and one-dimensional convolutional neural networks, also known as time delay neural networks (time delay neural networks), may be used to process one-dimensional data. The design concept of CNN is inspired by the visual neuroscience and mainly consists of a convolutional layer (convolutional layer) and a pooling layer (pooling layer). The convolutional layer can keep the spatial continuity of the image and can extract the local characteristics of the image. The pooling layer can adopt maximum pooling (max-pooling) or average pooling (mean-pooling), and the pooling layer can reduce the dimensionality of the middle hidden layer, reduce the computation of the next layers and provide rotation invariance.
In addition, the Long short-term Memory network (LSTM) uses an LSTM unit to replace a neuron in the RNN, and an input gate, an output gate and a forgetting gate are respectively added to input, output and forgetting past information to control the amount of the allowed information to pass through.
Furthermore, the prediction of the ship load power in the method is to determine the power distribution of the ship load in the three-phase circuit, the starting frequency and the running time length of the ship load according to the power load calculation book of the ship, analyze the step dynamic fluctuation characteristics of the ship load and the unbalance degree of the ship load in the three-phase circuit by combining the load demands of different running conditions, and construct a ship power load three-phase power prediction model based on a ship navigation plan.
Further, the marine photovoltaic power generation prediction model in the step S1 is constructed as follows:
firstly, the CNN model is used for extracting the spatial characteristics of the photovoltaic data, then the LSTM model is used for extracting the temporal characteristics of the photovoltaic data on the basis of the extracted spatial characteristics, and then the photovoltaic power is predicted.
Further, the method for constructing the photovoltaic power generation power prediction model for the ship in the step S1 comprises the following steps:
firstly, collecting potential influence factor data which can influence photovoltaic power generation power by a real ship, then preprocessing the data, modifying and filling the data, and then adopting a normalization formula
Carrying out normalization processing on the data;
dividing the processed data into a training set, a verification set and a test set, wherein the training set, the verification set and the test set are not crossed;
then photovoltaic data is input into a vector
Inputting the data into an upper CNN model to obtain the spatial characteristics of the photovoltaic data; the lower layer LSTM model stores time information about important photovoltaic data features extracted by the CNN;
and finally, obtaining the optimal lag time step, the layer number of each model in the mixed model and the parameters of each layer through a large amount of data training, and realizing the construction of the photovoltaic power generation power prediction model for the CNN-LSTM ship.
The normalization of the numerical type features can unify all the features into a roughly same numerical value interval, and is beneficial to constructing an original model.
Further, the influencing factors comprise external environment factors and navigation attitude data.
Furthermore, the data patch adopts a linear interpolation method and a weighted moving average method.
The linear interpolation is an interpolation mode in which an interpolation function is a first-order polynomial, and an interpolation error of the interpolation mode on an interpolation node is zero. Compared with other interpolation modes, such as parabolic interpolation, the linear interpolation has the characteristics of simplicity and convenience. The geometric meaning of linear interpolation is that the original function is approximately represented by a straight line passing through the points A and B in the overview chart. Linear interpolation can be used to approximate instead of primitive functions, or can be used to compute values that are not present in the table lookup process.
The weighted moving average method is a method of giving different weights to observed values, obtaining a moving average value according to the different weights, and determining a predicted value based on the final moving average value. The weighted moving average method is adopted because the recent observation value of the observation period has great influence on the predicted value, and the recent market change trend can be reflected better. Therefore, the observation values close to the forecast period are given larger weight values, the observation values far away from the forecast period are correspondingly given smaller weight values, and the action of each observation value on the forecast value is adjusted by different weight values, so that the forecast value can more approximately reflect the future development trend of the market.
In summary, and as shown in FIG. 1, the principle of the method is summarized as follows:
the invention relates to a source-load power prediction method applicable to a photovoltaic ship, wherein the source-load power prediction of a hybrid power system is mainly divided into two parts of predicting marine photovoltaic power generation power and ship load power, and the contents are as follows:
(1) And predicting the photovoltaic power generation power for the ship. The method is characterized by combining the photoelectric conversion nonlinear characteristic of the solar cell, considering the influence of ship speed, course, geographical position, ship swing angle, cloud layer and upper deck building shielding and deck surface temperature radiation on photovoltaic power generation power, analyzing the spatial correlation among factors and the correlation of single-factor time sequence data, researching and adopting a CNN-LSTM mixed deep neural network to construct a marine photovoltaic power generation power prediction model, and realizing the prediction of the marine photovoltaic power generation power.
(2) And predicting the load power of the ship. According to the 'electric load calculation book' of a large ship, power distribution of ship loads in a three-phase circuit, starting frequency and running time length of the ship loads are determined, and load requirements of different running working conditions are combined to analyze the step dynamic fluctuation characteristics of the ship loads and the unbalance degree of the ship loads in the three-phase circuit. And researching and constructing a ship power load three-phase power prediction model based on a ship navigation plan.
With the above contents combined, the implementation content of the source-load power prediction method applicable to the photovoltaic ship is as follows, firstly, a source-load power prediction model of a hybrid power system is researched and constructed by deeply analyzing the power fluctuation rule of a ship photovoltaic power generation system and a ship load; and then, constructing an electromagnetic transient simulation model of the hybrid power system, and revealing the influence mechanism of photovoltaic power generation power fluctuation, ship load high-power sudden change, diesel generator set start and stop, three-phase voltage normal unbalance and short-circuit fault on the hybrid power system.
Further, the source-load power prediction model of the hybrid power system is constructed by adopting a CNN-LSTM marine photovoltaic power generation power prediction model: the CNN-LSTM photovoltaic power prediction model to be constructed in the project is to extract the spatial features of photovoltaic data by using the CNN model, extract the temporal features of the photovoltaic data by using the LSTM model on the basis of the extracted spatial features, and predict the photovoltaic power, and the structure of the model is shown in FIG. 2. The specific implementation method comprises the following steps: firstly, collecting potential influence factor 0 (including external environment factors and navigation postures) data which can influence the photovoltaic power generation power by a real ship, then preprocessing the data, modifying and filling the data by adopting processing methods such as a linear interpolation method, a weighted moving average method and the like, and normalizing the data by adopting a normalization formula. The processed data is divided into a training set, a verification set and a test set, and are not crossed with each other. And inputting the photovoltaic data input vector into an upper CNN model to obtain the spatial characteristics of the photovoltaic data. The underlying LSTM model stores temporal information about important photovoltaic data features extracted by CNN. Finally, through a large amount of data training, obtaining the optimal lag time step, the number of layers of each model in the mixed model and the parameters of each layer, and realizing the construction of the photovoltaic power generation power prediction model for the CNN-LSTM ship;
further, the construction content of the ship load power prediction model provided by the invention is as follows: first, study on "electric load calculation book" of large ships, and classify the load use frequency (continuity, discontinuity, and sporadic); and then, according to the requirements of each navigation working condition of the ship on the load, determining the power distribution of the ship load in the three-phase circuit under different working conditions, and researching the three-phase power imbalance and the power fluctuation rule of the ship power grid. And finally, analyzing the time sequence of each navigation working condition in the whole navigation process according to the target ship navigation plan, and realizing the prediction of the ship load power.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrases "comprising a component of' 8230; \8230;" does not exclude the presence of another like element in a process, method, article, or apparatus that comprises the element.
The above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention, and all equivalent structures or equivalent processes performed by the present invention or directly or indirectly applied to other related technical fields are also included in the scope of the present invention.
Claims (10)
1. A 'source-load' power prediction method suitable for a photovoltaic ship is characterized by comprising the following steps:
s1, constructing a source-load power prediction model of a hybrid power system by analyzing power fluctuation rules of the power generation power of a marine photovoltaic power generation system and the power fluctuation rules of ship loads;
s2, establishing a prediction model according to the step S1, and constructing an electromagnetic transient simulation model of the hybrid power system, wherein the electromagnetic transient simulation model is used for predicting the influence mechanism of photovoltaic power generation power fluctuation, ship load high-power sudden change, diesel generator set start-stop, three-phase voltage normal unbalance and short-circuit fault on the hybrid power system;
and S3, combining a source-load power prediction model of the hybrid power system and an electromagnetic transient simulation model of the hybrid power system, and outputting a prediction result.
2. The source-load power prediction method suitable for the photovoltaic ship as claimed in claim 1, wherein in the method, the generation power prediction of the photovoltaic power generation system is to combine the photoelectric conversion nonlinear characteristic of the solar cell, consider the influence of the photovoltaic power generation, analyze the spatial correlation among factors and the correlation of single-factor time sequence data, and adopt a CNN-LSTM hybrid deep neural network to construct a marine photovoltaic power generation power prediction model.
3. The method of claim 2, wherein the photovoltaic power generation has an effect on one or more of ship speed, ship heading, geographic position, ship roll angle, cloud cover and upper deck building shading, and deck surface temperature radiation.
4. The 'source-load' power prediction method suitable for the photovoltaic ship according to claim 2, wherein in the method, the prediction of the ship load power is that according to a power load calculation book of the ship, the power distribution, the starting frequency and the running time length of the ship load in a three-phase circuit are determined, the load requirements of different running conditions are combined, the step dynamic fluctuation characteristics of the ship load and the unbalance degree of the ship load in the three-phase circuit are analyzed, and then a ship power load three-phase power prediction model is constructed based on a ship sailing plan.
5. The source-load power prediction method suitable for the photovoltaic ship as claimed in claim 1, wherein the electromagnetic transient simulation model of the hybrid power system includes data information of photovoltaic power generation power fluctuation, ship load high-power sudden change, diesel generator set start and stop, three-phase voltage normal imbalance and short-circuit fault.
6. The method for predicting the 'source-load' power suitable for the photovoltaic ship according to claim 4, wherein the step S1 of constructing the marine photovoltaic power generation prediction model comprises the following steps:
firstly, a CNN model is used for extracting the spatial characteristics of photovoltaic data, then on the basis of the extracted spatial characteristics, an LSTM model is used for extracting the temporal characteristics of the photovoltaic data, and then photovoltaic power prediction is carried out.
7. The source-load power prediction method suitable for the photovoltaic ship according to claim 6, characterized in that the marine photovoltaic power generation power prediction model construction method in the step S1 is as follows:
firstly, collecting potential influence factor data which can influence the photovoltaic power generation power by a real ship, preprocessing the data, modifying and filling the data, and then normalizing the data;
dividing the processed data into a training set, a verification set and a test set, wherein the training set, the verification set and the test set are not crossed;
then photovoltaic data is input into a vector
Inputting the data to an upper CNN model to obtain the spatial characteristics of the photovoltaic data; the lower LSTM model stores time information about important photovoltaic data features extracted through CNN;
and finally, obtaining the optimal lag time step, the number of layers of each model in the hybrid model and the parameters of each layer through a large amount of data training, and realizing the construction of the photovoltaic power generation power prediction model for the CNN-LSTM ship.
8. The method as claimed in claim 7, wherein the step data is normalized by a normalization formula
。
9. The method as claimed in claim 7, wherein the influencing factors include external environmental factors and navigation attitude data.
10. The method for predicting the 'source-load' power of the photovoltaic ship according to claim 7, wherein the data patch adopts a linear interpolation method and a weighted moving average processing method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211531908.6A CN115912352A (en) | 2022-12-01 | 2022-12-01 | Source-load power prediction method suitable for photovoltaic ship |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211531908.6A CN115912352A (en) | 2022-12-01 | 2022-12-01 | Source-load power prediction method suitable for photovoltaic ship |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115912352A true CN115912352A (en) | 2023-04-04 |
Family
ID=86493570
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211531908.6A Pending CN115912352A (en) | 2022-12-01 | 2022-12-01 | Source-load power prediction method suitable for photovoltaic ship |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115912352A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116738187A (en) * | 2023-08-08 | 2023-09-12 | 山东航宇游艇发展有限公司 | Ship gas power dynamic prediction method and system based on artificial intelligence |
-
2022
- 2022-12-01 CN CN202211531908.6A patent/CN115912352A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116738187A (en) * | 2023-08-08 | 2023-09-12 | 山东航宇游艇发展有限公司 | Ship gas power dynamic prediction method and system based on artificial intelligence |
| CN116738187B (en) * | 2023-08-08 | 2023-10-24 | 山东航宇游艇发展有限公司 | Ship gas power dynamic prediction method and system based on artificial intelligence |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wu et al. | Sustainability performance assessment of wind power coupling hydrogen storage projects using a hybrid evaluation technique based on interval type-2 fuzzy set | |
| Suganthi et al. | Applications of fuzzy logic in renewable energy systems–a review | |
| Wen et al. | Coordinated optimal energy management and voyage scheduling for all-electric ships based on predicted shore-side electricity price | |
| Quan et al. | Hybrid model for hourly forecast of photovoltaic and wind power | |
| CN104836259B (en) | Multi-energy resource ship micro-grid real-time energy optimal scheduling method | |
| Wen et al. | A joint photovoltaic-dependent navigation routing and energy storage system sizing scheme for more efficient all-electric ships | |
| Tao et al. | Joint optimization of wind turbine micrositing and cabling in an offshore wind farm | |
| Zhao et al. | Robust voltage control considering uncertainties of renewable energies and loads via improved generative adversarial network | |
| CN115912352A (en) | Source-load power prediction method suitable for photovoltaic ship | |
| Daraz | Optimized cascaded controller for frequency stabilization of marine microgrid system | |
| Chang | Comparison of three short term wind power forecasting methods | |
| Karabiber et al. | Short term PV power estimation by means of extreme learning machine and support vector machine | |
| Liu et al. | Review of grid stability assessment based on AI and a new concept of converter-dominated power system state of stability assessment | |
| Liu et al. | Optimal sitting and sizing of distributed generators based on a modified primal-dual interior point algorithm | |
| Liu et al. | Social implications of cyber-physical systems in electrical load forecasting | |
| Wei et al. | A novel energy management system for cruise ships including forecasting via LSTM | |
| Cheng et al. | Short-term load forecasting method for low voltage users based on deep belief neural network | |
| Ramirez-Rosado et al. | Powerful planning tools | |
| Berdonosov et al. | Analysis and development potential of predictive models for energy flows of autonomous hybrid energy systems | |
| Huang et al. | Energy management of integrated energy system in large ports | |
| Qian et al. | Deep learning-based short-term wind power prediction considering various factors | |
| Perera et al. | Optimum design of distributed energy hubs using hybrid surrogate models (HSM) | |
| Gaber et al. | Proposed Intelligent Energy Management Systems for Hybrid Electric Traction System | |
| Fei et al. | Uncertainty Analysis in the Power Grid Operation with Renewable Energy Generations | |
| Li et al. | Seasonal Forecasting Model of Data Center Net Load Based on LSTM-Attention Fusion Neural 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 |