Font Size: a A A

Study On The Impacts And Countermeasures Of Light Pollution At Sea On Ships’ Navigation Safety At Night

Posted on:2016-11-29Degree:DoctorType:Dissertation
Country:ChinaCandidate:J S ZhuFull Text:PDF
GTID:1222330470470032Subject:Traffic Information Engineering & Control
Abstract/Summary:PDF Full Text Request
In light of the accelerated process of industrialization and urbanization of human society, light pollution at sea has been deemed as a new type of pollution and becoming a severe threat to ships’ navigation safety at night, especially taking considerations of further development and utilization of marine resources.The mechanism of the light pollution at sea with regard to the ships’ navigation safety at night was systematically analyzed from the viewpoints of the concept of light pollution at sea, the principle of optical, colorimetry and visual performance, etc. Based on aforementioned analysis, the impact factors of light pollution at sea was proposed on ships’navigation safety of at night, which was taken as the main evaluation index for designing appropriate questionnaires; Statistical analysis was thereafter carried out on valid questionnaires by using SPSS and the correctness of the foregoing mechanism analysis was verified accordingly.Ship lights identification model based on BP neural network was established by taking the identifiability of ship lights as index for evaluating the level of light pollution at sea; In the aim of improving the efficiency of ship lights neural network identification model, the clonal selection algorithm was purposely improved; the optimization of directionality and efficiency was enhanced by introducing sensitivity analysis method to determine the optimal scale factor which was used to optimize the structure of BP neural network. The efficiency of the BP network model which was improved by clonal optimization algorithm was demonstrated by simulating the spot photos of ship lights at sea and recognition of ship lights.A designed method of ship lights system based on laser diode array was proposed with the purpose of improving identifiability of ship lights in the sea area where light pollution could occur. Based on the light-emitting theory of the conductor laser and the technical property requirements of ship lights, a design method of the ship laser lights was recommended by taking laser diode as light source. The light intensity, scope of arc light and color properties as claimed in the new-style ship lights was analyzed in the relevant simulation. The simulation results demonstrated that the new-style ship lights meet the requirements of all technical properties as stated in the COLREGs.Two countermeasures which can be used to avoid the impact of light pollution at sea on ships’ navigation safety of at night were summarized from legal point of view. First, in the 3rd article (general definition) of the current COLREGs, add the 14th provision ----"high brightness water area", and modify the title and content of the 3rd quarter (19th article) of the 2nd chapter in the COLREGs (the title will be modified to "Ship’s operation rules not in sight of one another"), so that the COLREGs can better guide the practice of collision avoidance and determine the obligation of anti-collision between ships; Second, based on the legal basis of analysis and determination of light pollution at sea, and the absence of the legislation in our country, this thesis put forward legislation ideas of the prevention and control of light pollution at sea.In conclusion, this thesis analyzed and verified the mechanism of the impact of light pollution at sea on ships’ navigation safety at night; Based on improved clonal algorithm, this thesis established the BP network identification model for ship lights, and achieved sound simulation results; From technical and legal aspects, countermeasures was recommended to prevent and govern light pollution at sea.
Keywords/Search Tags:Navigation safety of ships at night, Ship lights, Light pollution at sea, Regulations for preventing collisions, BP neural network
PDF Full Text Request
Related items