Flow online monitoring calculation method, monitor and monitoring system

Abstract

The invention provides a flow online monitoring calculation method, a monitor and a monitoring system. The method includes: step 1. data preprocessing and DCL model parameter initialization; step 2. corresponding to N H-ADCP cells, selecting a sample dimension; randomizing the initial population It is divided into 5 parts, each part is calculated corresponding to a model, and the output value of the model is obtained; the root mean square error of the output value and the measured value sample is used as the fitness function for optimization, and the smallest value is selected as the fitness function through fitness comparison. For the optimal index set of this iteration, record the optimal feature dimension set and parameter set; replace the global optimal fitness value with the selected one, and record the corresponding optimal feature dimension, calculation model and model parameters; If the accuracy and the maximum number of iterations are required, gradually expand the lower limit of the feature dimension selection and start the next iteration calculation; otherwise, output the optimal model and the corresponding model parameters, and give the final flow calculation value, and end the loop iteration; step 3. Parameter update .

Description

technical field [0001] The invention belongs to the field of hydrological measurement, and in particular relates to a calculation method for on-line flow monitoring, a monitoring instrument and a monitoring system. [0002] technical background [0003] In recent years, with the promulgation of systems such as the water resource management system and the system of river and lake chiefs, as well as the implementation of national key projects such as hydrological monitoring projects for small and medium-sized rivers, and the improvement of hydrological modernization capabilities, the timeliness and accuracy of hydrological monitoring elements have been raised. higher requirement. Among the many elements of hydrological monitoring, the most important and concerned is the monitoring of flow. The traditional flow monitoring mostly adopts the flow meter method, which not only has a large workload but also cannot achieve real-time monitoring. In recent years, with the development ...

AI Technical Summary

Problems solved by technology

[0006] (1) Calculation accuracy is not high: the current calculation methods are mostly based on mathematical statistics or the generalized curve formula of the vertical distribution of flow velocity as the core, but this method is only applicable to river sections with regular sections or relatively regular flow velocity distribution, and is applied under realistic conditions has great limitations, resulting in the calculation accuracy of this method is often not high

[0007] (2) Parameter selection problem: algorithms such as wavelet analysis and BP neural network cannot achieve the output of the optimal cell, and the calculation results often have large uncertainties, and cannot be adjusted with the increase of the number of samples or the change of the test section. Automatic adjustment of model parameters cannot be widely applied

Images