Operational Data Analysis And Energy Efficiency Optimization Method For Inland River Ship

With the continuous increase in the number of ships,the continuous improvement of transportation tasks,and the increasing frequency of economic activities,the probems in navigation efficiency and transportation benefits of inland river ship have more and more prominent.Ship operation data analysis and energy efficiency optimization are effective methods and main means to ensure ship transportation safety and enhance the ship operational bendfits.This paper studies operational data analysis and energy efficiency optimization method for inland river ship based on considering multiple influencing factors using the operation data of a whole transportation voyage.It proposes a series of methods including ship navigation status identification,daily water level of multiple stations prediction,and transportation segment division.Furthermore,the optimization methods and models for fuel consumption and transportation benefits on a whole voyage of the ship are constructed based on trajectory data repair,speed estimation,and fuel consumption rate estimation models.This research can provide methods and technical support for improving the operation data analysis,and energy efficiency optimization of inland river ship.The specific research contents and contributions mainly include the following aspects.(1)Collection and analysis of ship operation data in inland waters.Utilizing the online monitoring system of ship,the ship operation data of a whole voyage from Shanghai to Chongqing on the Yangtze River trunk line is collected.The sample data comprehensive preprocessing and ship navigation state recognition algorithms are put forward to extract the data of the ship’s normal navigation state,which provides the data basis for the research of this paper.(2)A method of multiple stations daily water level prediction for ship operation.A "divide and conquer" two-stage daily water level prediction method is proposed to deal with the daily water level data of 19 stations along the Yangtze River trunk line.First,the similarity measurement and hierarchical analysis-based clustering algorithm are performed for multiple stations’ water level data.Then,according to the periodic characteristics of the daily water level data,the prediction models for each cluster stations are constructed based on the Long Short-Term Memory neural network and Seasonal Autoregressive Integrated Moving Average method.This method improves the efficiency and accuracy of multiple stations’ daily water level prediction and provides support for the follow-up research of the thesis.(3)A method of inland river segment division.Aiming at the environment data including water level,water speed,wind speed and wind direction of a whole voyage,an algorithm of inland river segment division is proposed based on the density clustering analysis to obtain the characteristic variables of the segment.Also,a model based on the Long Short-Term Memory neural network is used to repair missing data along ship trajectory,and the segment division results are mapped to ship trajectory,which obtains the segments characteristic variable of the entire voyage.This method can reduce the influence of environment factors in inland river navigation on the operation data analysis and energy efficiency optimization,thus laying a foundation for the research on ship speed estimation,fuel consumption rate estimation,and energy efficiency optimization of the whole voyage.(4)A method for estimating the shipping speed and fuel consumption rate of inland river with the input of multiple characteristic variables.Based on the method of inland river segment division,a universal "modelling after the first division" method is proposed for operation data analysis.Based on the correlation analysis of characteristic variables,the model of ship speed estimation and fuel consumption rate estimation with multiple characteristic variables input is constructed based on General Regression Neural Network and Long Short-Term Memory neural network.The performance and advantages of the proposed models are verified through case studies and comparison with other methods,so as to provide support for the construction of ship fuel consumption and operational benefit optimization model for a whole voyage.(5)A method for ship fuel consumption and operational benefit optimization of a whole voyage considering multiple factors.Considering the impact of the complex navigable environment in inland river,the energy efficiency optimization problem of the whole voyage is subdivided into each segment.The optimization models,including the fuel consumption and operating efficiency for the whole voyage,are constructed by considering the constraints of the variable value interval with the lowest fuel consumption and the highest operating efficiency as the objectives,and the engine speed of each voyage as a decision variable.The obtained segment characteristic variables,the proposed speed estimation model and fuel consumption rate estimation model can support the objectives and constraints figuring out.The Improved Reduced Space Searching Algorithm is proposed through introducing the "synchronization-update" mechanism,which can solve the optimization models,to form the navigation scheme with the lowest fuel consumption and the highest operating efficiency of the whole voyage.The research results of this paper can be applied in the transportation of inland waters,to lift the efficiency of ship operation data analysis,and to improve the lowest fuel consumption and the most profitable navigation scheme,so as to promote the intelligent,green and sustainable development of inland shipping in China.