Study On Assessment Protocol And Prediction Model Of Urban Drainage Pipeline Conditions

Sewer pipeline is an important part of municipal infrastructure.It undertakes the key task of collecting and transporting waste water discharged by urban commerce,industry and housing as well as draining rainwater in time.Daily inspection and assessment of sewer pipeline is the key to effectively ensure its long-term healthy operation.The inspection and assessment of sewer pipelines around the world mainly consists of three parts: collect and obtain the internal image data of pipeline through the technology of detection robot equipment;based on the local pipeline assessment method,the structural and functional status of pipeline is assessed based on the collected pipeline image data;summarize the type,quantity,severity and other information of pipeline defects,and generate pipeline assessment report to guide subsequent repair works.Among them,the sewer pipeline condition assessment is the most critical step connecting the preceding and following,and different assessment methods have a decisive impact on the results of pipeline condition,but the research on the differences between different methods is insufficient.Based on the inspection and assessment data of sewer pipeline in the W City in South China,this paper studies the two mostly used assessment methods in the world,namely,the pipeline condition assessment protocol and the prediction model of pipeline condition.The main work contents and results are summarized as follows:(1)This paper systematically summarizes and introduces the development history and working principle of pipeline condition assessment protocols,and makes a comparative analysis research on four representative pipeline condition assessment protocols.The defect definition,defect weight,assessment parameters and Internal condition grades of the four protocols are qualitatively compared,and the equivalent mapping transformation between the protocols is realized;We also quantitatively compares the differences of the assessment results of the four protocols when evaluating the same pipelines.SRM-4assessment results are the most optimistic,and it is considered that 40% of the experimental pipelines belong to class 1;SPCCM is the most pessimistic,with 41% of the experimental pipelines considered to be class 4.The main reason for the difference in the assessment results of the four protocols is the design of the weight of defects and assessment parameters.The optimal design of assessment protocol at the regional level can adjust the defect weight and assessment parameters according to the local economic development,industrial distribution and land resource information.(2)A risk index method is proposed to study the correlation between drainage pipeline attributes and pipeline conditions based on descriptive statistical analysis Based on the pipeline information of Riverside District and City East District of the W City,the correlation between the most common pipeline information under the current inspection technologies(including pipeline type,diameter,length and material)and pipeline conditions,repair and maintenance needs and pipeline condition grade is studied by using the risk index method.The results show that the probability of health problems of sewage pipeline and rainwater pipeline is similar,but among the defective pipelines,the maintenance amount of sewage pipeline is 1.5 times that of rainwater pipeline.The smaller the pipe diameter,the higher the probability of health risk,and the more urgent the need for repair and maintenance;The pipeline with the length of 25m-35 m is in the best health condition;Concrete pipelines are less prone to structural defects,and plastic pipelines are less prone to functional failures.In addition,pipe diameter and pipe material are the most significant factors affecting the condition of drainage pipeline.(3)The prediction model of pipeline condition is a method to predict pipeline condition from pipeline detection data by fitting the relationship between pipeline attribute and pipeline condition.By summarizing the latest research results of current international prediction models of pipeline condition,this paper summarizes the pipeline detection data(mainly including pipeline attributes,operation conditions and environmental conditions)required by three types of prediction models(expert system,statistical model and machine learning model).According to the actual situation of China,the logistic regression analysis method is selected,combined with the research results of(2),and the prediction model of pipeline condition is developed based on the detection data of pipeline in the W City.The prediction model can correctly predict 54.86% of class2 pipelines and 51.40% of class 4 pipelines,but the prediction efficiency of class 1 and class 3 pipelines is not good.The experience of the model development is helpful to the construction of sewer pipeline data information database and the development of prediction model in the future.