Fuel consumption prediction method and device based on Bayesian network, equipment and medium

Abstract

The invention provides a fuel consumption prediction method and device based on a Bayesian network, equipment and a medium, and the method comprises the steps: obtaining first target navigation parameter information corresponding to the historical time of a ship and a corresponding first fuel consumption actual amount, and forming a data set; training a Bayesian network model by adopting a maximized posterior probability method based on the data set to obtain a trained Bayesian network model; inputting second target navigation parameter information corresponding to the current time of the ship into the trained Bayesian network model, and outputting a corresponding second fuel consumption prediction amount; according to the method, real-time accurate prediction of ship fuel consumption is realized through the Bayesian network model, small sample data processing is still effective, and under the condition that ship fuel data is limited or difficult to collect, limited samples can still be used for accurate prediction, so that powerful support is provided for ship management and operation.

Description

Technical Field

[0001] The present invention belongs to the technical field of ship performance prediction, and in particular relates to a fuel consumption prediction method, device, equipment and medium based on a Bayesian network. Background Art

[0002] Fuel cost is one of the main expenses of ship operation. Accurately predicting fuel consumption helps customers effectively control costs and optimize fuel procurement strategies. By accurately predicting fuel consumption, route planning can be optimized, unnecessary fuel consumption can be reduced, and operational efficiency can be improved.

[0003] However, existing fuel consumption prediction methods mainly rely on the experience of staff or require a large amount of historical data support, which leads to low accuracy in ship fuel consumption prediction. Summary of the invention

[0004] In a first aspect, an embodiment of the present invention provides a fuel consumption prediction method based on a Bayesian network, comprising: obtaining first target navigation parameter information corresponding to a ship at a historical time and a corresponding first actual fuel consumption amount to form a data set; training a Bayesian network model based on the data set using a maximum a posteriori probability method to obtain a trained Bayesian network model; inputting second target navigation parameter information corresponding to the ship at the current time into the trained Bayesian network model, and outputting the corresponding second fuel consumption prediction amount.

[0005] In some embodiments, after obtaining the first target navigation parameter information corresponding to the ship at the historical time, it also includes: when the first target navigation parameter information includes non-quantitative navigation parameters, quantizing the non-quantitative navigation parameter information.

[0006] In some embodiments, the obtaining of the first target navigation parameter information corresponding to the ship at the historical time includes: obtaining the first original navigation parameter information corresponding to the ship at the historical time; performing the following preprocessing on the first original navigation parameter information to obtain the first target navigation parameter information, and the preprocessing includes at least one of the following: data cleaning and feature engineering.

[0007] In some embodiments, the training of the Bayesian network model based on the data set using the maximization of posterior probability method to obtain a trained Bayesian network model includes: determining the prior probability of model parameters corresponding to the Bayesian network model based on expert knowledge; determining the likelihood function corresponding to the Bayesian network model according to the distribution characteristics of the data set, the likelihood function representing the distribution probability of the first historical target navigation parameter information and the first actual amount of fuel consumption observed under given model parameters; and obtaining the model parameters corresponding to the maximization of the posterior probability of the model parameters by maximizing the product of the prior distribution and the likelihood function.

[0008] In some embodiments, after constructing the data set, it also includes: segmenting the data set to obtain a training set and a test set; the method also includes the following item: obtaining multiple trained sub-Bayesian network models by cross-validation based on the training set, performing preset performance indicator evaluation on each sub-Bayesian network model, and determining the generalization performance of the trained Bayesian network model according to the preset performance indicator values ​​corresponding to the multiple sub-Bayesian network models; testing the trained Bayesian network model based on the test set, and determining the prediction performance of the trained Bayesian network model according to the error between the first fuel consumption prediction amount and the first fuel consumption actual amount corresponding to the test set.

[0009] In some embodiments, the method further includes: adjusting hyperparameters of the prior distribution or performing feature processing again according to the generalization performance or prediction performance.

[0010] In some embodiments, the method further comprises: regularly updating the trained Bayesian network model.

[0011] In the second aspect, an embodiment of the present invention provides a fuel consumption prediction device based on a Bayesian network, including: a historical data acquisition module, used to obtain the first target navigation parameter information corresponding to the ship at the historical time and the corresponding first fuel consumption actual amount, to form a data set; a prediction model training module, used to train the Bayesian network model based on the data set using the maximum a posteriori probability method to obtain a trained Bayesian network model; a fuel consumption prediction module, used to input the second target navigation parameter information corresponding to the ship at the current time into the trained Bayesian network model, and output the corresponding second fuel consumption prediction amount.

[0012] In a third aspect, an embodiment of the present invention provides an electronic device, characterized in that it includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus; the memory is used to store computer programs; the processor is used to implement the steps of the Bayesian network-based fuel consumption prediction method described in any one of the first aspects when executing the program stored in the memory.

[0013] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps of the Bayesian network-based fuel consumption prediction method as described in any one of the first aspects are implemented.

[0014] The beneficial effects brought by the present invention are as follows:

[0015] It can be seen from the above scheme that the embodiments of the present invention provide a fuel consumption prediction method, device, equipment and medium based on a Bayesian network, which realizes real-time and accurate prediction of ship fuel consumption through a Bayesian network model, and is still effective when processing small sample data. In the case of limited ship fuel data or difficult collection, limited samples can still be used for accurate prediction, providing strong support for ship management and operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic diagram of a flow chart of a fuel consumption prediction method based on a Bayesian network provided in an embodiment of the present invention;

[0017] Figure 2 for Figure 1 A detailed flow chart of step S102 in the illustrated embodiment;

[0018] Figure 3 A schematic flow chart of another fuel consumption prediction method based on Bayesian network provided in an embodiment of the present invention;

[0019] Figure 4 A schematic diagram of the structure of a fuel consumption prediction device based on a Bayesian network provided in an embodiment of the present invention;

[0020] Figure 5 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0021] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0022] Figure 1 A schematic diagram of a fuel consumption prediction method based on a Bayesian network provided by an embodiment of the present invention. Figure 1 As shown, the fuel consumption prediction method based on Bayesian network includes:

[0023] Step S101: Acquire first target navigation parameter information corresponding to a ship at a historical time and a corresponding first actual fuel consumption to form a data set.

[0024] Specifically, the target navigation parameter information includes the ship's navigation distance, navigation speed, ship type, load, weather conditions, etc. In this step, by collecting target navigation parameter information of different ships and different historical times and the corresponding actual fuel consumption, a data set including multiple samples is formed.

[0025] In some embodiments, the obtaining of the first target navigation parameter information corresponding to the ship at the historical time in step S101 includes: obtaining the first original navigation parameter information corresponding to the ship at the historical time; performing the following preprocessing on the first original navigation parameter information to obtain the first target navigation parameter information, and the preprocessing includes at least one of the following: data cleaning and feature engineering.

[0026] Specifically, when executing step S101, the original navigation parameter information is collected, and these original data need to be preprocessed to obtain the target navigation parameter information for modeling and analysis. Preprocessing usually includes two steps: data cleaning and feature engineering, where data cleaning refers to removing missing values, outliers and irrelevant data, and feature engineering refers to extracting and selecting feature parameters that affect fuel consumption.

[0027] In some embodiments, after the step S101 obtains the first target navigation parameter information corresponding to the ship at the historical time, it also includes: when the first target navigation parameter information includes non-quantitative navigation parameters, quantizing the non-quantitative navigation parameter information.

[0028] Specifically, the prediction of ship fuel consumption is affected by many factors, such as hull condition, navigation conditions, cargo loading, etc., all of which are uncertain. Quantifying these uncertain factors can convert subjective non-quantitative information into quantifiable data, which can be better integrated into the prediction model and improve the reliability and accuracy of the prediction.

[0029] Step S102: Based on the data set, a Bayesian network model is trained using a maximum a posteriori probability method to obtain a trained Bayesian network model.

[0030] Specifically, the problem definition was first clarified, that is, predicting the ship's fuel consumption as the prediction target; then the structure of the Bayesian network was determined according to the characteristics of the data set; then the sample data in the data set was used to estimate the conditional probability distribution between nodes in the Bayesian network; finally, the parameters of the Bayesian network model were obtained by maximizing the posterior probability, thereby obtaining a trained Bayesian network model.

[0031] Step S103: input the second target navigation parameter information corresponding to the ship at the current time into the trained Bayesian network model, and output the corresponding second fuel consumption prediction amount.

[0032] Specifically, the trained Bayesian network model can be deployed in the ship information system to achieve real-time prediction of fuel consumption.

[0033] In some embodiments, the method further comprises: regularly updating the trained Bayesian network model. Specifically, the prediction performance of the Bayesian network model can be regularly monitored, and new historical data can be obtained when necessary to continue updating the Bayesian network model.

[0034] The Bayesian network-based fuel consumption prediction method provided in this embodiment realizes real-time and accurate prediction of ship fuel consumption through the Bayesian network model, and is still effective when processing small sample data. In the case where ship fuel data is limited or difficult to collect, limited samples can still be used for accurate prediction, providing strong support for ship management and operation.

[0035] Based on the above embodiments, Figure 2 for Figure 1 A detailed flow chart of step S102 in the embodiment shown is as follows: Figure 2 As shown, step S102 includes the following steps:

[0036] Step S201: Determine the prior probability of model parameters corresponding to the Bayesian network model based on expert knowledge.

[0037] Specifically, the calculation formula corresponding to the Bayesian network model is: y=f(w,x), where y represents the predicted fuel consumption, w represents the model parameters of the Bayesian network model, and x represents the target navigation parameter information; in this step, experts can select the prior probability distribution of the model parameters based on past historical data, which can be recorded as: P(w).

[0038] Step S202: determining a likelihood function corresponding to the Bayesian network model according to the distribution characteristics of the data set, wherein the likelihood function represents the distribution probability of the first historical target navigation parameter information and the first actual amount of fuel consumption observed under given model parameters.

[0039] Specifically, a likelihood function can be defined based on the data distribution characteristics in the data set. The corresponding calculation formula is: P(x, y|w), which represents the probability of observing data x and y under given model parameters w, assuming that they obey a normal distribution.

[0040] Step S203: Obtain model parameters corresponding to the maximum posterior probability of model parameters by maximizing the product of the prior distribution and the likelihood function.

[0041] Specifically, based on the Bayesian theorem combined with the prior distribution and the likelihood function, the posterior probability of the model parameters P(w|x,y)∝P(x,yw)·P(w) is calculated. By maximizing the product of the prior distribution and the likelihood function, the posterior probability of the model parameters is maximized, and the model parameters corresponding to the maximum posterior probability of the model parameters are obtained, thereby obtaining a trained Bayesian network model.

[0042] In some embodiments, after the step S101 forms a data set, it also includes: segmenting the data set to obtain a training set and a test set; the method also includes the following item: obtaining multiple trained sub-Bayesian network models by cross-validation based on the training set, performing preset performance indicator evaluation on each sub-Bayesian network model, and determining the generalization performance of the trained Bayesian network model according to the preset performance indicator values ​​corresponding to the multiple sub-Bayesian network models; testing the trained Bayesian network model based on the test set, and determining the prediction performance of the trained Bayesian network model according to the error between the first fuel consumption prediction amount and the first fuel consumption actual amount corresponding to the test set.

[0043] Specifically, the data set is first split into a training set and a test set. The training set is used to train the Bayesian network model, while the test set is used to evaluate the performance of the model. The next steps include the following:

[0044] Cross-validation: Divide the training set into k subsets, each subset is used as a validation set in turn, and the remaining k-1 subsets are used as sub-training sets; train a sub-Bayesian network model on each sub-training set, and evaluate its preset performance indicators, such as prediction accuracy, mean square error, etc., on the corresponding validation set; determine the generalization performance of the trained Bayesian network model by calculating the preset performance indicator values ​​(such as the average value) of the k sub-Bayesian network models. It should be noted that when selecting the trained Bayesian network model for subsequent applications, the sub-Bayesian network model with the best preset performance indicator can be selected from the k sub-Bayesian network models, or the Bayesian network model trained on the entire training set can be selected.

[0045] Prediction and evaluation: Input the historical target navigation parameter information in the test set into the trained Bayesian network model and output the corresponding historical fuel consumption prediction; compare the predicted historical fuel consumption with the actual historical fuel consumption to evaluate the prediction performance of the Bayesian network model.

[0046] In some embodiments, the method further comprises: adjusting the hyperparameters of the prior distribution or performing feature processing again according to the generalization performance or prediction performance. Specifically, adjusting the hyperparameters of the prior distribution or further optimizing feature selection according to the performance of the Bayesian network model.

[0047] On the basis of the foregoing embodiments, the accuracy of the Bayesian network model prediction is further improved by integrating the experts' prior knowledge and experience into the training process of the Bayesian network model. The performance of the trained Bayesian network model is also comprehensively evaluated, including the performance on the training set, generalization ability, and prediction performance on the test set, thereby ensuring the accuracy and robustness of the model.

[0048] In order to further understand the embodiments of the present invention, Figure 3 A schematic diagram of a flow chart of another fuel consumption prediction method based on a Bayesian network provided in an embodiment of the present invention, such as Figure 3 As shown, the following steps are included:

[0049] (1) Problem definition: Predicting the fuel consumption of a ship is the prediction objective.

[0050] (2) Historical data collection: Collect the historical fuel records of the ship and related historical navigation parameters, such as sailing distance, sailing speed, ship type, deadweight, and weather conditions, to form a data set.

[0051] (3) Data preprocessing: data cleaning to remove missing values, outliers, and irrelevant data; feature engineering to extract and select feature parameters that affect fuel; data quantization to convert non-quantitative information into quantitative data; data segmentation to divide the data set into a training set and a test set.

[0052] (4) Establishing a Bayesian network model: Experts select the prior distribution of model parameters based on historical data and define the likelihood function of fuel based on the data distribution characteristics in the data set, assuming that it obeys a normal distribution.

[0053] (5) Bayesian network model training: Use Bayes’ theorem combined with the prior distribution and likelihood function to calculate the posterior distribution of model parameters; use the posterior probability of the maximum model parameter to determine the model parameters.

[0054] (6) Verification of the generalization performance of the Bayesian network model: The cross-validation method is used to evaluate the generalization ability of the Bayesian network model, including performance indicators such as prediction accuracy and mean square error.

[0055] (7) Prediction and evaluation of the Bayesian network model: Use the trained Bayesian network model to predict the fuel consumption of the test set; compare the predicted fuel consumption with the actual fuel consumption to evaluate the prediction performance of the Bayesian network model.

[0056] (8) Optimization of the Bayesian network model: Adjust the hyperparameters of the prior distribution according to the performance of the Bayesian network model to further optimize feature selection.

[0057] (9) Deployment and application of Bayesian network model: Integrate the trained Bayesian network model into the ship information system to predict fuel consumption in real time; and regularly monitor the model performance and update the model when necessary.

[0058] In summary, this embodiment provides a fuel consumption prediction method based on a Bayesian network, which has the following advantages over existing methods: First, this embodiment can effectively process and quantify the uncertainty of fuel, thereby providing a more reliable prediction; second, it supports the integration of experts' prior knowledge and experience into this embodiment, thereby improving the accuracy of the prediction; third, this embodiment is still effective when processing small sample data. When it is difficult to collect ship fuel data, limited samples can still be used for accurate prediction.

[0059] Figure 4 A schematic diagram of the structure of a fuel consumption prediction device based on a Bayesian network provided by an embodiment of the present invention, such as Figure 4 As shown, the fuel consumption prediction device based on the Bayesian network includes:

[0060] The historical data acquisition module 401 is used to acquire the first target navigation parameter information and the corresponding first actual fuel consumption of the ship at a historical time to form a data set;

[0061] A prediction model training module 402 is used to train a Bayesian network model based on the data set using a maximum a posteriori probability method to obtain a trained Bayesian network model;

[0062] The fuel consumption prediction module 403 is used to input the second target navigation parameter information corresponding to the ship at the current time into the trained Bayesian network model, and output the corresponding second fuel consumption prediction amount.

[0063] In some embodiments, the historical data acquisition module 401 is further used to:

[0064] In the case where the first target navigation parameter information includes non-quantitative navigation parameters, the non-quantitative navigation parameter information is quantized.

[0065] In some embodiments, the historical data acquisition module 401 is specifically used to:

[0066] Obtaining the first original navigation parameter information corresponding to the ship at the historical time;

[0067] The first original navigation parameter information is preprocessed as follows to obtain the first target navigation parameter information, wherein the preprocessing includes at least one of the following: data cleaning and feature engineering.

[0068] In some embodiments, the prediction model training module 402 is specifically used to:

[0069] Determine the prior probability of model parameters corresponding to the Bayesian network model based on expert knowledge;

[0070] Determining a likelihood function corresponding to the Bayesian network model according to the distribution characteristics of the data set, wherein the likelihood function represents the distribution probability of the first historical target navigation parameter information and the first actual amount of fuel consumption observed under given model parameters;

[0071] The model parameters corresponding to the maximum posterior probability of the model parameters are obtained by maximizing the product of the prior distribution and the likelihood function.

[0072] In some embodiments, the historical data acquisition module 401 is further used to:

[0073] Segmenting the data set to obtain a training set and a test set;

[0074] The prediction model training module 402 is further used for one of the following:

[0075] Based on the training set, a plurality of trained sub-Bayesian network models are obtained by cross-validation, a preset performance indicator is evaluated for each sub-Bayesian network model, and the generalization performance of the trained Bayesian network model is determined according to the preset performance indicator values ​​corresponding to the plurality of sub-Bayesian network models;

[0076] The trained Bayesian network model is tested based on the test set, and the prediction performance of the trained Bayesian network model is determined according to the error between the first fuel consumption prediction amount and the first fuel consumption actual amount corresponding to the test set.

[0077] In some embodiments, the prediction model training module 402 is further used to:

[0078] According to the generalization performance or prediction performance, the hyperparameters of the prior distribution are adjusted or feature processing is performed again.

[0079] In some embodiments, the prediction model training module 402 is further used to:

[0080] The trained Bayesian network model is updated regularly.

[0081] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process and corresponding beneficial effects of the fuel consumption prediction device based on the Bayesian network described above can refer to the corresponding process in the aforementioned method example and will not be repeated here.

[0082] like Figure 5 As shown, an embodiment of the present invention provides an electronic device, including a processor 501, a communication interface 502, a memory 503 and a communication bus 504, wherein the processor 501, the communication interface 502, and the memory 503 communicate with each other through the communication bus 504.

[0083] Memory 503, used for storing computer programs;

[0084] In one embodiment of the present invention, the processor 501 is used to implement the steps of the Bayesian network-based fuel consumption prediction method provided by any of the aforementioned method embodiments when executing the program stored in the memory 503.

[0085] The implementation principle and technical effect of the electronic device provided by the embodiment of the present invention are similar to those of the above embodiment and will not be described in detail here.

[0086] The above-mentioned memory 503 can be an electronic memory such as a flash memory, an EEPROM (electrically erasable programmable read-only memory), an EPROM, a hard disk or a ROM. The memory 503 has a storage space for program codes for executing any method steps in the above-mentioned method. For example, the storage space for program codes may include various program codes for implementing various steps in the above method respectively. These program codes can be read from or written into one or more computer program products. These computer program products include program code carriers such as hard disks, compact disks (CDs), memory cards or floppy disks. Such computer program products are usually portable or fixed storage units. The storage unit may have a storage segment or storage space arranged similarly to the memory 503 in the above-mentioned electronic device. The program code can be compressed, for example, in an appropriate form. Generally, the storage unit includes a program for executing the method steps according to an embodiment of the present invention, that is, a code that can be read by a processor such as 501, which, when run by an electronic device, causes the electronic device to execute various steps in the method described above.

[0087] The embodiment of the present invention further provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the fuel consumption prediction method based on the Bayesian network are implemented.

[0088] The computer-readable storage medium may be included in the device / apparatus described in the above embodiment; or it may exist independently without being assembled into the device / apparatus. The above computer-readable storage medium carries one or more programs, and when the above one or more programs are executed, the method according to the embodiment of the present invention is implemented.

[0089] According to an embodiment of the present invention, the computer-readable storage medium may be a non-volatile computer-readable storage medium, for example, may include but is not limited to: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, the computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, apparatus, or device.

[0090] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0091] The above are preferred embodiments of the present invention. It should be pointed out that, for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A fuel consumption prediction method based on Bayesian network, characterized in that: include: Acquire the first target navigation parameter information and the corresponding first actual fuel consumption of the ship at the historical time to form a data set; Based on the data set, a Bayesian network model is trained using a maximum a posteriori probability method to obtain a trained Bayesian network model; The second target navigation parameter information corresponding to the ship at the current time is input into the trained Bayesian network model, and the corresponding second fuel consumption prediction amount is output.

2. The method according to claim 1, characterized in that After obtaining the first target navigation parameter information corresponding to the ship at the historical time, the method further includes: In the case where the first target navigation parameter information includes non-quantitative navigation parameters, the non-quantitative navigation parameter information is quantized.

3. The method according to claim 2, characterized in that The obtaining of the first target navigation parameter information corresponding to the ship at the historical time includes: Obtaining the first original navigation parameter information corresponding to the ship at the historical time; The first original navigation parameter information is preprocessed as follows to obtain the first target navigation parameter information, wherein the preprocessing includes at least one of the following: data cleaning and feature engineering.

4. The method according to any one of claims 1 to 3, characterized in that: The training of the Bayesian network model based on the data set using the maximum a posteriori probability method to obtain a trained Bayesian network model includes: Determine the prior probability of model parameters corresponding to the Bayesian network model based on expert knowledge; Determining a likelihood function corresponding to the Bayesian network model according to the distribution characteristics of the data set, wherein the likelihood function represents the distribution probability of the first historical target navigation parameter information and the first actual amount of fuel consumption observed under given model parameters; The model parameters corresponding to the maximization of the posterior probability of the model parameters are obtained by maximizing the product of the prior distribution and the likelihood function.

5. The method according to claim 4, characterized in that After the data set is constructed, the following is also included: Segmenting the data set to obtain a training set and a test set; The method further comprises one of the following: Based on the training set, a plurality of trained sub-Bayesian network models are obtained by cross-validation, a preset performance indicator is evaluated for each sub-Bayesian network model, and the generalization performance of the trained Bayesian network model is determined according to the preset performance indicator values ​​corresponding to the plurality of sub-Bayesian network models; The trained Bayesian network model is tested based on the test set, and the prediction performance of the trained Bayesian network model is determined according to the error between the first fuel consumption prediction amount and the first fuel consumption actual amount corresponding to the test set.

6. The method according to claim 5, characterized in that The method further comprises: According to the generalization performance or prediction performance, the hyperparameters of the prior distribution are adjusted or feature processing is performed again.

7. The method according to any one of claims 1 to 3, characterized in that: The method further comprises: The trained Bayesian network model is updated regularly.

8. A fuel consumption prediction device based on Bayesian network, characterized in that: include: A historical data acquisition module, used to acquire first target navigation parameter information and first actual fuel consumption corresponding to the ship at a historical time to form a data set; A prediction model training module is used to train a Bayesian network model based on the data set using a maximum a posteriori probability method to obtain a trained Bayesian network model; The fuel consumption prediction module is used to input the second target navigation parameter information corresponding to the ship at the current time into the trained Bayesian network model and output the corresponding second fuel consumption prediction amount.

9. An electronic device, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus; Memory, used to store computer programs; The processor is used to implement the steps of the Bayesian network-based fuel consumption prediction method described in any one of claims 1 to 7 when executing the program stored in the memory.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the Bayesian network-based fuel consumption prediction method as described in any one of claims 1 to 7 are implemented.

Images