Research On Identification And Substitution Of Abnormal Value Of Flow Rate Data Under Different Oxygen Enrichment Modes

In the process of fish culture,water velocity is an important environmental factor affecting fish growth,development and reproduction.The Acoustic Doppler Velocimetry(ADV)is a practical device for measuring the flow velocity.It has the advantages of high measurement accuracy,fast response and easy operation.It is widely used in field exploration and hydrodynamic simulation experiments in laboratory.Aeration device is an important dissolved oxygen supplement device in pond culture.However,when it is started,silt suspension and bubbles will appear in the aquaculture water flow environment,which will affect the stability of the probe of the current meter and result in abnormal values in the monitoring of the flow rate.Therefore,it is necessary to post-process the monitoring data of the flow rate to improve the accuracy of the data.Based on the 3D robust phase space method,this paper designs and develops a post-processing software for velocity data of Doppler current meter on the QT platform,which is used to identify and replace outliers in velocity measurement results.The flow rate data post-processing software constructed in this study was used to verify the processing effect of simulated flow rate data with a known number of outliers.Flow rate monitoring experiments were carried out in ponds No.4 and No.7 of the Chinese Mitten crab breeding ponds in Baoshan District of Shanghai and No.1 of Chongming District when the micro-pore aeration device and the water-pushing aeration device were put into operation.The constructed flow rate data post-processing software was used to detect and replace the outliers of the monitoring results.The effect of velocity data post-processing software on detecting and processing anomalies in different aeration modes is analyzed.The main research results are as follows:(1)The post-processing software of flow rate data shows that the correct detection rate of outliers is more than 95%.The kurtosis coefficients of flow rate data with different outliers show a downward trend after processing,and the characteristic number of average asymmetry degree tends to 0,and the curve becomes smooth and close to the original data.Both the artificially inserted extreme outliers and the approximate true outliers are effectively detected.It indicates that the post-processing software can effectively identify and replace outliers in velocity data.(2)In the micropore aeration,the surface and bottom flow rate data processing results of short-term sampling showed that the proportion of outliers was 23.14% and25.61%,respectively.After the flow rate data post-processing software detection and processing,the kurtosis coefficient decreased by 72.72% and 61.47%,respectively,and the maximum and minimum values in the data set were close to the average value.The standard deviation tends to decrease,and the average skewness tends to 0.The extreme peak value in the flow velocity data is effectively identified and accurately replaced,and the processed data is smoother.In the long-term flow rate sampling results,the outlier processing results of the two groups of flow rate data show that the flow fluctuation range of the first group is small,accounting for 3.40% of the outlier,while the flow fluctuation range of the second group is large,and the number of outliers is significantly increased,accounting for 12.57%.Before the post-processing software detection and processing of velocity data,the velocity fluctuation frequency was high and the amplitude was large.After the processing,the extreme peak value of the two monitoring velocity was effectively identified and replaced.The standard deviation showed a decreasing trend,the kurtosis coefficient decreased by 38.69% and 43.61%,respectively,and the average skew tended to 0.This indicates that the outliers in both short-and long-term flow rate data can be effectively detected.(3)After the micropore aeration and water pushing aeration devices were opened at the same time,extreme peaks appeared in the surface and bottom sampling data,with the proportion of abnormal values being 14.34% and 17.73%,respectively.After the post-processing software of the flow rate data,the kurtosis coefficient decreased by56.02% and 87.26%,the standard deviation decreased by 11.4% on average,and the average skewness all tended to 0.The effective reduction of data dispersion indicates that the post-processing software of velocity data can effectively remove outliers from velocity monitoring data and improve the accuracy of velocity monitoring data on the basis of the discrete trend of original data.After comprehensive analysis,the post-processing software of velocity data of Doppler current meter based on 3D robust phase space method and QT construction is suitable for the processing of outliers of monitoring velocity data under different aeration modes,in order to provide accurate velocity data for accurate and quantitative research on hydrodynamics of pond aquaculture system.