River and lake integrated ecological dredging monitoring method and system based on hull

Abstract

The present invention discloses a hull-based integrated ecological dredging monitoring method and system for rivers and lakes, and belongs to the technical field of intelligent control systems applied to municipal administration. It includes: obtaining real-time data of impurity removal status and obtaining impurity removal abnormal data; determining impurity removal abnormal points based on impurity removal abnormal data and obtaining adjustment abnormal data; determining adjustment abnormal points based on adjustment abnormal data; obtaining time series data of pressing status and obtaining pressing abnormal data, determining pressing abnormal points based on pressing abnormal data and performing abnormal processing; displaying and recording impurity removal abnormal points, adjustment abnormal points and pressing abnormal points in a visual form. The present invention realizes all-round monitoring while reducing the amount of data processing. The point where equipment failure occurs can be directly obtained through data association between models, that is, the alarm link can be directly displayed on the system platform, which is convenient for staff to make accurate judgments and give response strategies in a timely manner.

Description

Technical Field

[0001] The intelligent control system of the present invention is applied in the technical field of municipal engineering, and specifically relates to a hull-based river and lake integrated ecological dredging monitoring method and system. Background Art

[0002] In order to solve the inconvenience of traditional sludge treatment technology, hull dredging methods came into being. A hull is built at a designated location in a river or lake, and the equipment required for dredging is integrated on the hull, such as a suction device, a sludge removal device, a sludge regulating tank, a sludge press, etc. The above-mentioned integrated large-scale equipment overcomes the problems of existing technologies such as large disturbances caused by long-distance dredging and transportation to land for treatment, insufficient treatment, and shortage of land resources. It is also suitable for river and lake areas with thick silt layers and many obstacles. For areas with thin silt layers or complex water environments, it may be necessary to combine other dredging methods for comprehensive management.

[0003] Therefore, the above-mentioned large equipment is built on rivers and lakes, and many integrated devices are required. During operation, it is necessary to effectively supervise the collaborative work of multiple systems and multiple devices to ensure the stable operation of the system and maximize the dredging effect. In order to ensure the comprehensiveness and timeliness of monitoring, corresponding sensors will be installed in each link. However, each link on the hull is not only equipped with one set of equipment, but generally multiple sets of settings are set to work simultaneously or staggered. Therefore, there are many types and quantities of equipment, and a large demand for sensors. Although real-time data monitoring can be achieved, the amount of data is large, so the data processing volume is large, the storage volume is large, the operation is relatively complex, and the background operation performance on the hull is high.

[0004] In addition, it is also necessary to consider that the silt treatment on the hull is different from that on land. It is necessary to consider in advance the impact of weather conditions, power supply and other factors on system operation and equipment to avoid the inability to perform timely maintenance and repairs on the system due to environmental factors. Summary of the invention

[0005] In view of the problems in the related technology, the present invention proposes a hull-based integrated river and lake ecological dredging monitoring method and system to overcome the above-mentioned technical problems existing in the existing related technology.

[0006] To this end, the specific technical solution adopted by the present invention is as follows: a river and lake integrated ecological dredging monitoring method based on a hull, the ecological dredging monitoring method comprising the following steps:

[0007] Obtaining real-time data on the impurity removal status based on the sludge impurity removal device , the real-time data of the impurity removal status Inputting into the abnormal data analysis model to obtain the impurity removal abnormal data; creating a sludge impurity removal alarm model, determining the impurity removal abnormal points in the impurity removal link based on the impurity removal abnormal data and performing abnormal processing;

[0008] Obtain the time series data of the regulation status of the sludge regulating tank , the adjustment state timing data Inputting into the abnormal data analysis model to obtain abnormal adjustment data; creating a sludge adjustment alarm model, determining the abnormal adjustment point in the adjustment link or impurity removal link based on the abnormal adjustment data and performing abnormal processing;

[0009] Obtaining the time series data of the squeezing status of the sludge press , the squeezing state time series data Input into the abnormal data analysis model to obtain the pressing abnormal data; create a sludge pressing alarm model, determine the pressing abnormal points in the pressing link, adjustment link or impurity removal link based on the pressing abnormal data and perform abnormal processing;

[0010] The removal abnormal points, adjustment abnormal points and pressing abnormal points are displayed and recorded in a visual form.

[0011] In a further embodiment, the ecological dredging monitoring method further comprises the following steps:

[0012] The real-time data of the impurity removal status obtained each time , Adjustment status time series data and squeeze state time series data The data are stored in a database and defined as historical state data, and the historical state data are divided into a training set and a test set in proportion;

[0013] Constructing a data prediction model, and continuously optimizing the data prediction model using a training set and a test set to obtain an optimized data prediction model;

[0014] Select the corresponding real-time status data in the form of time series data as the input of the data prediction model, and obtain the status prediction data for the future time period based on the optimized data prediction model; input the status prediction data into the abnormal data analysis model to obtain the abnormal prediction data; and input the abnormal prediction data into the corresponding alarm model according to the type of the abnormal prediction data to obtain the predicted status alarm data.

[0015] In a further embodiment, the real-time data of the impurity removal status It is expressed in the following form: ; In the formula, It indicates the cleaning working parameters pre-set according to the cleaning requirements. Indicates the parameters of the impurity removal work The vibrating screen obtained under the operating environment The real-time weight on Indicates the parameters of the impurity removal work The vibrating screen obtained under the operating environment The real-time vibration frequency on the data; correspondingly, the impurity-removed abnormal data at least includes: abnormal weight data and abnormal vibration data;

[0016] The adjustment state time series data It is expressed in the following form: ; In the formula, The working parameters of the regulating pool are pre-set according to the regulation requirements. To adjust the working parameters of the pool Suspension water level in the working environment, To adjust the working parameters of the pool Suspension flow rate under working environment, To adjust the working parameters of the pool correspondingly, the abnormal adjustment data at least includes: abnormal suspension water level data, abnormal suspension flow rate data and abnormal sludge contamination degree data;

[0017] The squeezing state time series data It is expressed in the following form: ; In the formula Indicates the preset working parameters of the filter press. Indicates filter press working parameters The squeezing pressure obtained under the operating environment, Working parameters of filter press The feed liquid level obtained under the operating environment, Indicates filter press operating parameters The filtrate flow rate is obtained under the operating environment, and the abnormal pressing data at least includes: abnormal feed liquid level data, abnormal pressure data, and abnormal filtrate flow rate data.

[0018] In a further embodiment, the analysis process of the abnormal data analysis model is as follows:

[0019] Divide the state time series data under the same working parameters into current monitoring data according to the time frame and historical monitoring data, and the historical monitoring data is formed into a window data sequence ,in, Represents the historical time frame; the current monitoring data is calculated using the following formula The rate of change :

[0020] ;

[0021] In the formula, for and If the time interval If it has been identified as abnormal monitoring data, The calculation is repeated until ; If the window data sequence If there is no data or all the data are abnormal monitoring data, the current monitoring data is directly determined. Normal monitoring data;

[0022] Based on the rate of change Calculate the normalization coefficient : , where is the upper threshold, is the lower threshold;

[0023] Establish corresponding evaluation criteria according to the type of monitoring data, and normalize the coefficients based on the corresponding evaluation criteria. Evaluation: If the evaluation is normal monitoring data, the abnormal coefficient of the normal monitoring data is set to 0; if the evaluation is abnormal monitoring data, As the abnormal coefficient of the abnormal monitoring data.

[0024] In a further embodiment, the impurity removal link at least includes: an impurity removal input link, a sludge impurity removal link and an impurity removal output link; correspondingly, the analysis process of the sludge impurity removal alarm model is as follows:

[0025] Step 101: If abnormal weight data is obtained, obtain the corresponding weight abnormality coefficient ,like It is judged that there is a blockage in the impurity removal input link; if Then find out whether there is abnormal vibration data in the same time frame: if there is abnormal vibration data, execute step 102; if there is no abnormal vibration data, execute step 103;

[0026] Step 102, determining whether a vibration power source of the sludge impurity removal process fails;

[0027] Step 103, finding out whether there is abnormal suspension liquid water level data in the same time frame: if there is abnormal suspension liquid water level data, executing step 104; if there is no abnormal suspension liquid water level data, executing step 105;

[0028] Step 104: Obtaining the abnormal coefficient of the suspension water level ,like It is judged that the impurity removal output link is blocked; if Then execute step 105;

[0029] Step 105: It is determined that the vibration screen of the sludge impurity removal process is blocked.

[0030] In a further embodiment, the adjustment link at least includes: an adjustment input link, an additive preparation link, a stirring link and an adjustment output link; the analysis process of the sludge adjustment alarm model is as follows:

[0031] If abnormal suspension water level data is obtained, the corresponding suspension water level abnormality coefficient is obtained. ,like It is judged that the adjustment input link is blocked; if Determine whether there is a blockage in the regulation output link;

[0032] If abnormal suspension flow rate data is obtained, the corresponding suspension flow rate abnormality coefficient is obtained. ,like If It is judged that the stirring speed of the stirring link is too fast;

[0033] If abnormal silt concentration data is obtained, the corresponding silt concentration abnormality coefficient is obtained. ,Right now , perform the following steps:

[0034] Step 201, finding whether there is abnormal suspension flow rate data in the current same time frame: if there is no abnormal suspension flow rate data, executing step 202; if there is abnormal suspension flow rate data, executing step 203;

[0035] Step 202: determine whether there is a problem in the additive preparation process;

[0036] Step 203: Obtaining the abnormal coefficient of suspension flow rate ,like If Then execute step 202.

[0037] In a further embodiment, the squeezing process at least includes: a pressure process, a feeding process, a squeezing process, a sewage discharge process and a mud cake output process; the analysis process of the sludge squeezing alarm model is as follows:

[0038] If abnormal feed level data is obtained, the corresponding feed level abnormality coefficient is obtained. , and perform the following steps: If , then execute step 301; if , then execute step 302;

[0039] Step 301, find out whether there is abnormal suspension water level data in the same time frame. If there is no abnormal suspension water level data, execute step 303; if there is abnormal suspension water level data, obtain the suspension water level abnormality coefficient , Then execute step 302, if , then execute step 303; if , then analyze whether the adjustment input link is blocked in combination with the sludge adjustment alarm model, if not, execute step 303;

[0040] Step 302: determine whether the impurity removal output link is blocked;

[0041] Step 303, determining whether a feed link is blocked;

[0042] If abnormal pressure data is obtained, obtain the pressure anomaly coefficient ,like , it is judged that the pressure is lower than the pressure requirement;

[0043] If the evaluation obtains abnormal filtrate flow data, obtain the filtrate flow abnormality coefficient ,Right now , then analyze whether there is other abnormal squeeze data at present, if yes, execute step 401, if no, execute step 402;

[0044] Step 401: If the data is abnormal feed level data, it is determined that the impurity removal output link is blocked; if the data is abnormal pressure data, it is determined that the pressure is lower than the pressure requirement;

[0045] Step 402: It is determined that the filtration time is too short.

[0046] In a further embodiment, the construction and training process of the data prediction model is as follows:

[0047] Building a data prediction model: ; In the formula, represents the constant value of the initial learner, is the target value of the nth training sample, N is the number of training samples, is the loss function, c is the constant when the loss function is minimized;

[0048] For each sample The negative gradient is calculated using the following formula:

[0049] ;

[0050] In the formula, For the establishment of A regression tree for the sample The residual fit of For the The predicted value of a regression tree, is the learner constant;

[0051] Based on data Fitting to get The leaf node area corresponding to the regression tree is ,in, , For the The number of leaf nodes in a regression tree, is the input quantity; for each leaf node Calculate the best fit value :

[0052] ;

[0053] The optimized data prediction model is expressed as: ; In the formula, is the indicator function.

[0054] In a further embodiment, the state prediction data at least includes: impurity removal state prediction data, adjustment state prediction data and pressing state prediction data; correspondingly, the abnormality prediction data at least includes: impurity removal abnormality prediction data, adjustment abnormality prediction data and pressing abnormality prediction data;

[0055] The alarm model at least includes: a sludge impurity removal alarm model, a sludge regulation alarm model and a sludge pressing alarm model.

[0056] The integrated ecological dredging monitoring system for rivers and lakes based on a hull is used to implement the integrated ecological dredging monitoring method for rivers and lakes as described above, comprising: a hull, wherein the hull is pre-arranged with a suction device, a sludge removal device, a sludge regulating tank, a sludge press, and a mud cake conveying device according to the needs;

[0057] It also includes: a plurality of weight sensors and vibration sensors, which are correspondingly installed in the sludge removal device;

[0058] A plurality of first water level sensors, flow rate sensors and concentration sensors are correspondingly arranged at designated positions of the sludge regulating tank;

[0059] A plurality of second water level sensors, pressure sensors and flow sensors are correspondingly arranged in the sludge press;

[0060] And the first module is configured to obtain real-time data of the impurity removal status based on the sludge impurity removal device , the real-time data of the impurity removal status Inputting into the abnormal data analysis model to obtain the impurity removal abnormal data; creating a sludge impurity removal alarm model, determining the impurity removal abnormal points in the impurity removal link based on the impurity removal abnormal data and performing abnormal processing;

[0061] The second module is configured to obtain the time series data of the regulation status of the sludge regulating tank , the adjustment state timing data Inputting into the abnormal data analysis model to obtain abnormal adjustment data; creating a sludge adjustment alarm model, determining the abnormal adjustment point in the adjustment link or impurity removal link based on the abnormal adjustment data and performing abnormal processing;

[0062] The third module is configured to obtain the time series data of the squeezing state of the sludge press , the squeezing state time series data Input into the abnormal data analysis model to obtain the pressing abnormal data; create a sludge pressing alarm model, determine the pressing abnormal points in the pressing link, adjustment link or impurity removal link based on the pressing abnormal data and perform abnormal processing;

[0063] The fourth module is configured to display and record the removal abnormal points, adjustment abnormal points and compression abnormal points in a visual form.

[0064] Beneficial effects of the present invention: Considering that the working environment on the hull is different from that on the ground, the uncontrollable factors increase, such as weather, water level, equipment operating environment and monitoring cycle, etc. Therefore, the monitoring and prediction of a series of equipment on the hull becomes particularly important, and the difficulty of monitoring increases relatively.

[0065] Therefore, the present invention discloses a hull-based integrated river and lake ecological dredging monitoring method and system, which reduces the amount of data processing while achieving all-round monitoring through abnormal data analysis models, sludge removal alarm models, sludge regulation alarm models, and sludge pressing alarm models. The point where equipment failure occurs can be directly obtained through data association between models, that is, the alarm link is directly displayed on the system platform, which is convenient for staff to make accurate judgments and give response strategies in a timely manner.

[0066] Different from the requirement of installing a sensor in each link of the prior art, the present invention only needs to install corresponding sensors in key links, and can obtain meshed correlation data by combining sludge impurity removal alarm model, sludge adjustment alarm model, and sludge pressing alarm model. The fault point can be directly determined through the meshed correlation data. Therefore, the demand for sensors is reduced in terms of hardware requirements, thereby reducing the sensor failure rate. In terms of software processing, the hardware operation requirements for data collection and transmission are reduced, and the data processing efficiency is improved.

[0067] In addition, the sensors in the prior art are all independent individuals, so the collected data may be abnormal data caused by equipment failure, or abnormal data caused by sensor failure itself, which requires further manual judgment to determine. The present application, on the other hand, again uses meshed correlation data to circumvent this problem, and reduces the possibility of misjudgment due to sensor failure in the prior art through relevant data logic. In addition, a data prediction model is set up to monitor abnormal data in the future, that is, to perform maintenance in advance through predicted abnormal data. It is more suitable for use on the hull, avoiding the problem of being unable to board the ship for maintenance due to weather conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 It is a flow chart of the ecological dredging monitoring method of Example 1.

[0069] Figure 2 It is an analysis flow chart of the sludge impurity removal alarm model of Example 1.

[0070] Figure 3 It is an analysis flow chart of the sludge regulation alarm model of Example 1.

[0071] Figure 4 It is an analysis flow chart of the sludge squeezing alarm model of Example 1. DETAILED DESCRIPTION

[0072] To further illustrate each embodiment, the present invention provides drawings, which are part of the disclosure of the present invention and are mainly used to illustrate the embodiments. They can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these contents, ordinary technicians in this field should be able to understand other possible implementation methods and advantages of the present invention.

[0073] Example 1

[0074] This embodiment provides an integrated ecological dredging monitoring method for rivers and lakes based on a ship hull. The method is also applicable to ecological dredging monitoring on land. This embodiment is described using the use of a ship hull as an example.

[0075] like Figure 1 As shown, the ecological dredging monitoring method includes the following steps:

[0076] Obtaining real-time data on the impurity removal status based on the sludge impurity removal device , the real-time data of the impurity removal status Inputting into the abnormal data analysis model to obtain the impurity removal abnormal data; creating a sludge impurity removal alarm model, determining the impurity removal abnormal points in the impurity removal link based on the impurity removal abnormal data and performing abnormal processing;

[0077] Obtain the time series data of the regulation status of the sludge regulating tank , the adjustment state timing data Inputting into the abnormal data analysis model to obtain abnormal adjustment data; creating a sludge adjustment alarm model, determining the abnormal adjustment point in the adjustment link or impurity removal link based on the abnormal adjustment data and performing abnormal processing;

[0078] Obtaining the time series data of the squeezing status of the sludge press , the squeezing state time series data The abnormal pressing data is input into the abnormal data analysis model to obtain the abnormal pressing data; a sludge pressing alarm model is created, and based on the abnormal pressing data, the abnormal pressing points are determined in the pressing link, the adjustment link or the impurity removal link and the abnormal processing is performed.

[0079] Finally, the removal abnormal points, adjustment abnormal points and pressing abnormal points are displayed and recorded in a visual form.

[0080] In a further embodiment, the real-time data of the cleaning status It is expressed in the following form: ; In the formula, It indicates the cleaning working parameters pre-set according to the cleaning requirements. Indicates the parameters of the impurity removal work The vibrating screen obtained under the operating environment The real-time weight on Indicates the parameters of the impurity removal work The vibrating screen obtained under the operating environment correspondingly, the impurity-removed abnormal data at least includes: abnormal weight data and abnormal vibration data.

[0081] It is worth mentioning that the impurity removal working parameter p is limited in this embodiment, that is, the parameter setting is different under different requirements. Further, the impurity removal working parameter p can include: filter mesh aperture, filter mesh load-bearing weight, vibration frequency and vibration duration, etc. according to the degree of sludge contamination.

[0082] Taking the silt situation in different areas of the same river or lake as an example, the silt in area A is more polluted, so it has relatively more pollutants and shellfish products, so the particle sizes of pollutants vary, so the aperture levels of the filter will be different, and the vibration time and vibration amplitude requirements will be higher during impurity removal. The silt in area B is less polluted, and the particle size of its pollutants is relatively small, so the aperture requirement of the filter is a fine pore size. Therefore, the settings of the impurity removal working parameter p in area A and area B are different. Therefore, when doing data analysis, the real-time data of the same type of equipment parameters are compared and analyzed to increase the reliability of the data analysis. At the same time, in order to ensure the authenticity of the subsequent abnormal data analysis, reduce errors, and improve alarm accuracy,

[0083] Correspondingly, adjust the state timing data It is expressed in the following form: ; In the formula, The working parameters of the regulating pool are pre-set according to the regulation requirements. The suspension water level range, stirring speed, additive concentration, etc. can be set according to actual needs. To adjust the working parameters of the pool Suspension water level in working environment , To adjust the working parameters of the pool Suspension flow rate under working environment, To adjust the working parameters of the pool correspondingly, the abnormal adjustment data at least includes: abnormal suspension water level data, abnormal suspension flow rate data and abnormal sludge pollution degree data.

[0084] Squeeze status time series data It is expressed in the following form: ; In the formula Indicates the preset filter press working parameters. Indicates filter press working parameters The squeezing pressure obtained under the operating environment, Working parameters of filter press The feed liquid level obtained under the operating environment, Indicates filter press operating parameters The filtrate flow rate is obtained under the operating environment, and the abnormal pressing data at least includes: abnormal feed liquid level data, abnormal pressure data, and abnormal filtrate flow rate data.

[0085] Combined with the above introduction to the data collection type of this embodiment, it can be seen that the data collected in this embodiment is only: vibrating screen Real-time weight, vibrating screen The real-time vibration frequency, suspension water level, suspension flow rate, sludge concentration, pressing pressure, feed level and filtrate flow rate are collected. The above data are the key data for each link. These key data correspond to the input and output ends, and can form a network of related data. The fault point can be directly determined through the network of related data. Therefore, the number of sensors required in hardware requirements is reduced, thereby reducing the sensor failure rate and the delay of abnormal processing.

[0086] At the same time, it also solves the possibility of misjudgment due to sensor failure in the prior art, because the data of the present application is related through the sludge / sludge processing line, and the data correlation is linked. The abnormal point obtained through multiple data is more accurate.

[0087] In terms of software processing, the operating environment and configuration requirements for data collection and transmission are reduced, and data processing efficiency is improved.

[0088] That is, it is different from the requirement of installing a sensor in each link of the prior art. Taking the impurity removal link as an example, the prior art generally installs a feed flow sensor and a discharge flow sensor at the input end and the output end to judge whether there is a blockage, but the present application does not need to install such sensors, and can be detected by the vibrating screen. Real-time weight, vibrating screen The real-time vibration frequency is obtained by removing the vibrating screen In addition to the relevant status of the vibration motor, the working status of the input and output ends can also be obtained.

[0089] The specific method is as follows: first, create an abnormal data analysis model. The analysis process of the abnormal data analysis model in this embodiment is as follows:

[0090] Divide the state time series data under the same working parameters into current monitoring data according to the time frame and historical monitoring data, and the historical monitoring data is formed into a window data sequence ,in, Represents a historical time frame;

[0091] Use the following formula to calculate the current monitoring data The rate of change :

[0092] ;

[0093] In the formula, for and If the time interval If it has been identified as abnormal monitoring data, The calculation is repeated until ; If the window data sequence If there is no data or all the data are abnormal monitoring data, the current monitoring data will be directly determined. Normal monitoring data;

[0094] Based on the rate of change Calculate the normalization coefficient : , where is the upper threshold, is the lower threshold;

[0095] Establish corresponding evaluation criteria according to the type of monitoring data (the evaluation criteria can be found below), and normalize the normalization coefficient based on the corresponding evaluation criteria. Evaluation: If the evaluation is normal monitoring data, the abnormal coefficient of the normal monitoring data is set to 0; if the evaluation is abnormal monitoring data, As the abnormal coefficient of the abnormal monitoring data.

[0096] Among them, the upper threshold and lower threshold The method of obtaining is as follows: Get a set of historical monitoring data series , n is the number of historical monitoring data in this group, according to the formula Calculate the corresponding rate of change sequence Based on the change rate sequence Perform frequency analysis to obtain the sequence frequency density function and the cumulative probability distribution function (Probability distribution fitting).

[0097] Set the confidence level to ,by and Distribution is upper threshold and lower threshold .

[0098] Furthermore, the impurity removal process at least includes: an impurity removal input process, a sludge impurity removal process and an impurity removal output process; Figure 2 ,The analysis process of the sludge impurity removal alarm model is as follows:

[0099] Step 101: If abnormal weight data is obtained, obtain the corresponding weight abnormality coefficient ,like It is judged that there is a blockage in the impurity removal input link; if Then find out whether there is abnormal vibration data in the same time frame: if there is abnormal vibration data, execute step 102; if there is no abnormal vibration data, execute step 103;

[0100] Step 102, determining whether a vibration power source of the sludge impurity removal process fails;

[0101] Step 103, finding out whether there is abnormal suspension liquid water level data in the same time frame: if there is abnormal suspension liquid water level data, executing step 104; if there is no abnormal suspension liquid water level data, executing step 105;

[0102] Step 104: Obtaining the abnormal coefficient of the suspension water level ,like It is judged that the impurity removal output link is blocked; if Then execute step 105;

[0103] Step 105: It is determined that the vibration screen of the sludge impurity removal process is blocked.

[0104] The above logical relationship can be used to directly locate the fault point, which can be displayed in a visual form in the background running program. Correspondingly, it also reflects the correlation of data. If one of the sensors fails, the abnormal point can also be determined through the above logical relationship.

[0105] The adjustment process at least includes: adjustment input process, additive preparation process, stirring process and adjustment output process. Figure 3 , the analysis process of the sludge regulation alarm model is as follows:

[0106] If abnormal suspension water level data is obtained, the corresponding suspension water level abnormality coefficient is obtained. ,like It is judged that the adjustment input link is blocked; if Determine whether there is a blockage in the regulation output link;

[0107] If abnormal suspension flow rate data is obtained, the corresponding suspension flow rate abnormality coefficient is obtained. ,like If It is judged that the stirring speed of the stirring link is too fast;

[0108] If abnormal silt concentration data is obtained, the corresponding silt concentration abnormality coefficient is obtained. ,Right now , perform the following steps:

[0109] Step 201, finding whether there is abnormal suspension flow rate data in the current same time frame: if there is no abnormal suspension flow rate data, executing step 202; if there is abnormal suspension flow rate data, executing step 203;

[0110] Step 202: determine whether there is a problem in the additive preparation process;

[0111] Step 203: Obtaining the abnormal coefficient of suspension flow rate ,like If Then execute step 202.

[0112] In a further embodiment, the squeezing process at least includes: a pressure application process, a feeding process, a squeezing process, a sewage discharge process and a mud cake output process. Figure 4 As shown in the figure, the analysis process of the sludge squeezing alarm model is as follows:

[0113] If abnormal feed level data is obtained, the corresponding feed level abnormality coefficient is obtained. , and perform the following steps: If , then execute step 301; if , then execute step 302;

[0114] Step 301, find out whether there is abnormal suspension water level data in the same time frame. If there is no abnormal suspension water level data, execute step 303; if there is abnormal suspension water level data, obtain the suspension water level abnormality coefficient , Then execute step 302, if , then execute step 303; if , then analyze whether the adjustment input link is blocked in combination with the sludge adjustment alarm model, if not, execute step 303;

[0115] Step 302: determine whether the impurity removal output link is blocked;

[0116] Step 303, determining whether a feed link is blocked;

[0117] If abnormal pressure data is obtained, obtain the pressure anomaly coefficient ,like , then it is judged that the pressure is lower than the pressure requirement;

[0118] If the evaluation obtains abnormal filtrate flow data, obtain the filtrate flow abnormality coefficient ,Right now , then analyze whether there is other abnormal squeeze data at present, if yes, execute step 401, if no, execute step 402;

[0119] Step 401: If the data is abnormal feed level data, it is determined that the impurity removal output link is blocked; if the data is abnormal pressure data, it is determined that the pressure is lower than the pressure requirement;

[0120] Step 402: It is determined that the filtration time is too short.

[0121] In summary, the specific process of establishing corresponding evaluation criteria according to the monitoring data type is as follows:

[0122] The evaluation criteria for weight data are in the following form: Normalization coefficient of weight monitoring data , it is evaluated as normal monitoring data and its abnormal coefficient is assigned to be 0; otherwise, it is evaluated as abnormal monitoring data. Customize it as its abnormal coefficient. is the minimum weight evaluation threshold of the set impurity removal working parameter p, It is the maximum weight evaluation threshold of the set impurity removal working parameter p.

[0123] The evaluation criteria for vibration frequency data are in the following form: if the normalized coefficient of the frequency monitoring data is less than the vibration threshold, it is evaluated as abnormal vibration data; otherwise, it is normal monitoring data and its abnormal coefficient is assigned to be 0.

[0124] The evaluation criteria for suspension water level data are in the following form: Normalization coefficient of suspension water level monitoring data , it is evaluated as normal monitoring data and its abnormal coefficient is assigned to be 0; otherwise, it is evaluated as abnormal monitoring data. Customize it as its abnormal coefficient. is the minimum suspension water level evaluation threshold of the regulating tank working parameter b, It is the maximum suspension water level evaluation threshold of the regulating tank working parameter b.

[0125] The evaluation criteria for suspension flow rate data are in the following form: Normalization coefficient of suspension flow rate monitoring data , it is evaluated as normal monitoring data and its abnormal coefficient is assigned to be 0; otherwise, it is evaluated as abnormal monitoring data. Customize it as its abnormal coefficient. is the minimum suspension flow rate evaluation threshold of the regulating tank working parameter b, It is the maximum suspension flow rate evaluation threshold of the regulating tank working parameter b.

[0126] The evaluation criteria for sludge pollution monitoring data are in the following form: Normalization coefficient of sludge pollution monitoring data , it is evaluated as abnormal monitoring data; otherwise, it is normal monitoring data and its abnormal coefficient is assigned as 0. It is the evaluation threshold value of sludge pollution degree of regulating pond working parameter b.

[0127] The evaluation criteria for feed level monitoring data are in the following form: Normalization coefficient of feed level monitoring data , it is evaluated as normal monitoring data and its abnormal coefficient is assigned to be 0; otherwise, it is evaluated as abnormal monitoring data. Customize it as its abnormal coefficient. Among them, is the minimum feed level evaluation threshold of the filter press operating parameter d, It is the maximum feed liquid level evaluation threshold of the filter press operating parameter d.

[0128] The pressure monitoring data evaluation criteria are in the following form: Normalization coefficient of pressure monitoring data , it is evaluated as abnormal monitoring data; otherwise, it is normal monitoring data and its abnormal coefficient is assigned as 0. is the pressure evaluation threshold of the filter press working parameter d.

[0129] The evaluation criteria for filtrate flow monitoring data are in the following form: Normalization coefficient of filtrate flow monitoring data , it is evaluated as abnormal monitoring data; otherwise, it is normal monitoring data and its abnormal coefficient is assigned as 0. is the filtrate flow evaluation threshold of the filter press operating parameter d.

[0130] Considering that maintenance on the hull is relatively difficult, it is necessary to land on the hull for maintenance. If the maintenance personnel cannot perform normal maintenance due to weather conditions, and an alarm message has been issued, the processing chain will be shelved, which may cause more serious damage to the equipment. Therefore, in order to avoid such a situation in advance, this embodiment also includes the following steps:

[0131] The real-time data of the impurity removal status obtained each time , Adjustment status time series data and squeeze state time series data The data are stored in a database and defined as historical state data, and the historical state data are divided into a training set and a test set in proportion;

[0132] Constructing a data prediction model, and continuously optimizing the data prediction model using a training set and a test set to obtain an optimized data prediction model;

[0133] Select the corresponding real-time status data in the form of time series data as the input of the data prediction model, and obtain the status prediction data for the future time period based on the optimized data prediction model; input the status prediction data into the abnormal data analysis model to obtain abnormal prediction data; input the abnormal prediction data into the corresponding alarm model according to the type of the abnormal prediction data to obtain the predicted status alarm data. That is, repeat the analysis process of the alarm model above.

[0134] That is, the monitoring data of the future time period is predicted through historical data, and the upcoming alarm is predicted through the predicted monitoring data, especially before the bad weather comes, the abnormal points corresponding to the maintenance alarm are pre-checked or maintained to ensure that maintenance can be carried out in advance if it is impossible to log in in time. This not only ensures the service life of the equipment, but also improves the production capacity.

[0135] In order to achieve the above functions, the construction and training process of the data prediction model is as follows:

[0136] Building a data prediction model: ; In the formula, represents the constant value of the initial learner, is the target value of the nth training sample, N is the number of training samples, is the loss function, c is the constant when the loss function is minimized;

[0137] For each sample The negative gradient is calculated using the following formula:

[0138] ;

[0139] In the formula, For the establishment of A regression tree for the sample The residual fit of For the The predicted value of a regression tree, is the learner constant;

[0140] Data-based Fitting to get The leaf node area corresponding to the regression tree is ,in, , For the The number of leaf nodes in a regression tree, is the input quantity; for each leaf node Calculate the best fit value :

[0141] ;

[0142] The optimized data prediction model is expressed as: ; In the formula, is the indicator function.

[0143] Further, the state prediction data at least includes: impurity removal state prediction data, adjustment state prediction data and pressing state prediction data; correspondingly, the abnormality prediction data at least includes: impurity removal abnormality prediction data, adjustment abnormality prediction data and pressing abnormality prediction data;

[0144] The alarm model at least includes: a sludge impurity removal alarm model, a sludge regulation alarm model and a sludge pressing alarm model.

[0145] Example 2

[0146] This embodiment discloses a river-lake integrated ecological dredging monitoring system, which is used to implement the river-lake integrated ecological dredging monitoring method described in Embodiment 1, including: a hull, wherein the hull is pre-arranged with a winch suction device, a sludge impurity removal device, a sludge regulating tank, a sludge press, and a mud cake conveying device according to demand;

[0147] It also includes: a plurality of weight sensors and vibration sensors, which are correspondingly installed in the sludge removal device;

[0148] A plurality of first water level sensors, flow rate sensors and concentration sensors are correspondingly arranged at designated positions of the sludge regulating tank;

[0149] A plurality of second water level sensors, pressure sensors and flow sensors are correspondingly arranged in the sludge press;

[0150] And the first module is configured to obtain real-time data of the impurity removal status based on the sludge impurity removal device , the real-time data of the impurity removal status Inputting into the abnormal data analysis model to obtain the impurity removal abnormal data; creating a sludge impurity removal alarm model, determining the impurity removal abnormal points in the impurity removal link based on the impurity removal abnormal data and performing abnormal processing;

[0151] The second module is configured to obtain the time series data of the regulation status of the sludge regulating tank , the adjustment state timing data Inputting into the abnormal data analysis model to obtain abnormal adjustment data; creating a sludge adjustment alarm model, determining the abnormal adjustment point in the adjustment link or impurity removal link based on the abnormal adjustment data and performing abnormal processing;

[0152] The third module is configured to obtain the time series data of the squeezing state of the sludge press , the squeezing state time series data Input into the abnormal data analysis model to obtain the pressing abnormal data; create a sludge pressing alarm model, determine the pressing abnormal points in the pressing link, adjustment link or impurity removal link based on the pressing abnormal data and perform abnormal processing;

[0153] The fourth module is configured to display and record the removal abnormal points, adjustment abnormal points and compression abnormal points in a visual form.

Claims

1. A hull-based integrated river and lake ecological dredging monitoring method, characterized in that: The ecological dredging monitoring method comprises the following steps: Obtaining real-time data on the impurity removal status based on the sludge impurity removal device , the real-time data of the impurity removal status Inputting into the abnormal data analysis model to obtain the impurity removal abnormal data; creating a sludge impurity removal alarm model, determining the impurity removal abnormal points in the impurity removal link based on the impurity removal abnormal data and performing abnormal processing; Obtain the time series data of the regulation status of the sludge regulating tank , the adjustment state timing data Inputting into the abnormal data analysis model to obtain abnormal adjustment data; creating a sludge adjustment alarm model, determining the abnormal adjustment point in the adjustment link or impurity removal link based on the abnormal adjustment data and performing abnormal processing; Obtaining the time series data of the squeezing status of the sludge press , the squeezing state time series data Input into the abnormal data analysis model to obtain the pressing abnormal data; create a sludge pressing alarm model, determine the pressing abnormal points in the pressing link, adjustment link or impurity removal link based on the pressing abnormal data and perform abnormal processing; The impurity removal abnormal points, adjustment abnormal points and squeezing abnormal points are displayed and recorded in a visual form; the analysis process of the abnormal data analysis model is as follows: Divide the state time series data under the same working parameters into current monitoring data according to the time frame and historical monitoring data, and the historical monitoring data is formed into a window data sequence ,in, Represents the historical time frame; the current monitoring data is calculated using the following formula The rate of change : ； In the formula, for and If the time interval If it has been identified as abnormal monitoring data, The calculation is repeated until ; If the window data sequence If there is no data or all the data are abnormal monitoring data, the current monitoring data is directly determined. Normal monitoring data; Based on the rate of change Calculate the normalization coefficient : , where is the upper threshold, is the lower threshold; Establish corresponding evaluation criteria according to the type of monitoring data, and normalize the coefficients based on the corresponding evaluation criteria. Evaluation: If the evaluation is normal monitoring data, the abnormal coefficient of the normal monitoring data is set to 0; if the evaluation is abnormal monitoring data, As the abnormal coefficient of the abnormal monitoring data; The analysis process of the sludge removal alarm model is as follows: Step 101: If abnormal weight data is obtained, obtain the corresponding weight abnormality coefficient ,like It is judged that there is a blockage in the impurity removal input link; if Then find out whether there is abnormal vibration data in the same time frame: if there is abnormal vibration data, execute step 102; if there is no abnormal vibration data, execute step 103; wherein, is the minimum weight evaluation threshold of the set impurity removal working parameter p, is the maximum weight evaluation threshold of the set impurity removal working parameter p; Step 102, determining whether a vibration power source of the sludge impurity removal process fails; Step 103, finding out whether there is abnormal suspension liquid water level data in the same time frame: if there is abnormal suspension liquid water level data, executing step 104; if there is no abnormal suspension liquid water level data, executing step 105; Step 104: Obtaining the abnormal coefficient of the suspension water level ,like It is judged that the impurity removal output link is blocked; if Then execute step 105; is the minimum suspension water level evaluation threshold of the regulating tank working parameter b, is the maximum suspension water level evaluation threshold of the regulating tank working parameter b; Step 105: determining that the vibration screen of the sludge impurity removal process is clogged; The analysis process of the sludge adjustment alarm model is as follows: If abnormal suspension water level data is obtained, the corresponding suspension water level abnormality coefficient is obtained. ,like It is judged that the adjustment input link is blocked; if Determine whether there is a blockage in the regulation output link; is the minimum feed level evaluation threshold of the filter press operating parameter d, is the maximum feed level evaluation threshold of the filter press operating parameter d; If abnormal suspension flow rate data is obtained, the corresponding suspension flow rate abnormality coefficient is obtained. ,like If It is judged that the stirring speed of the stirring link is too fast; is the minimum suspension flow rate evaluation threshold of the regulating tank working parameter b, is the maximum suspension flow rate evaluation threshold of the regulating tank working parameter b; If abnormal silt concentration data is obtained, the corresponding silt concentration abnormality coefficient is obtained. ,Right now , is the sludge pollution evaluation threshold of the regulating tank working parameter b; then perform the following steps: Step 201, finding whether there is abnormal suspension flow rate data in the current same time frame: if there is no abnormal suspension flow rate data, executing step 202; if there is abnormal suspension flow rate data, executing step 203; Step 202, determining whether there is a problem in the additive preparation process; Step 203: Obtaining the abnormal coefficient of suspension flow rate ,like If Then execute step 202; The analysis process of the sludge squeezing alarm model is as follows: If abnormal feed level data is obtained, the corresponding feed level abnormality coefficient is obtained. , and perform the following steps: If , then execute step 301; if , then execute step 302, wherein, is the minimum feed level evaluation threshold of the filter press operating parameter d, is the maximum feed level evaluation threshold of the filter press operating parameter d; Step 301, find out whether there is abnormal suspension water level data in the same time frame. If there is no abnormal suspension water level data, execute step 303; if there is abnormal suspension water level data, obtain the suspension water level abnormality coefficient , Then execute step 302, if , then execute step 303; if , then analyze whether the adjustment input link is blocked in combination with the sludge adjustment alarm model, if not, execute step 303; Step 302: determine whether the impurity removal output link is blocked; Step 303, determining whether a feed link is blocked; If abnormal pressure data is obtained, obtain the pressure anomaly coefficient ,like , then it is judged that the pressure is lower than the pressure requirement, is the pressure evaluation threshold of the filter press working parameter d; If the evaluation obtains abnormal filtrate flow data, obtain the filtrate flow abnormality coefficient ,Right now , is the filtrate flow evaluation threshold of the filter press working parameter d; then analyze whether there is other abnormal pressing data at present, if yes, execute step 401, if no, execute step 402; Step 401: If the data is abnormal feed level data, it is determined that the impurity removal output link is blocked; if the data is abnormal pressure data, it is determined that the pressure is lower than the pressure requirement; Step 402: It is determined that the filtration time is too short.

2. The integrated river and lake ecological dredging monitoring method based on a hull according to claim 1 is characterized in that: The ecological dredging monitoring method further comprises the following steps: The real-time data of the impurity removal status obtained each time , Adjustment status time series data and squeeze state time series data The data are stored in a database and defined as historical state data, and the historical state data are divided into a training set and a test set in proportion; Constructing a data prediction model, and continuously optimizing the data prediction model using a training set and a test set to obtain an optimized data prediction model; Select the corresponding real-time status data in the form of time series data as the input of the data prediction model, and obtain the status prediction data for the future time period based on the optimized data prediction model; input the status prediction data into the abnormal data analysis model to obtain the abnormal prediction data; and input the abnormal prediction data into the corresponding alarm model according to the type of the abnormal prediction data to obtain the predicted status alarm data.

3. The integrated river and lake ecological dredging monitoring method based on a hull according to claim 1 is characterized in that: The real-time data of the impurity removal status It is expressed in the following form: ; In the formula, It indicates the cleaning working parameters pre-set according to the cleaning requirements. Indicates the parameters of the impurity removal work The vibrating screen obtained under the operating environment The real-time weight on Indicates the parameters of the impurity removal work The vibrating screen obtained under the operating environment Real-time vibration frequency on; Correspondingly, the removed abnormal data at least includes: abnormal weight data and abnormal vibration data; The adjustment state time series data It is expressed in the following form: ; In the formula, The working parameters of the regulating pool are pre-set according to the regulation requirements. To adjust the working parameters of the pool Suspension water level in the working environment, To adjust the working parameters of the pool Suspension flow rate under working environment, To adjust the working parameters of the pool correspondingly, the abnormal adjustment data at least includes: abnormal suspension water level data, abnormal suspension flow rate data and abnormal sludge contamination degree data; The squeezing state time series data It is expressed in the following form: ; In the formula Indicates the preset working parameters of the filter press. Indicates filter press working parameters The squeezing pressure obtained under the operating environment, Working parameters of filter press The feed liquid level obtained under the operating environment, Indicates filter press operating parameters The filtrate flow rate is obtained under the operating environment, and the abnormal pressing data at least includes: abnormal feed liquid level data, abnormal pressure data, and abnormal filtrate flow rate data.

4. The integrated river and lake ecological dredging monitoring method based on a hull according to claim 1 is characterized in that: The impurity removal link at least includes: an impurity removal input link, a sludge impurity removal link and an impurity removal output link.

5. The integrated river and lake ecological dredging monitoring method based on a hull according to claim 1 is characterized in that: The regulating link at least includes: an regulating input link, an additive preparation link, a stirring link and a regulating output link.

6. The integrated river and lake ecological dredging monitoring method based on a hull according to claim 1 is characterized in that: The squeezing process at least includes: a pressure application process, a feeding process, a squeezing process, a sewage discharge process and a mud cake output process.

7. The integrated river and lake ecological dredging monitoring method based on a hull according to claim 2 is characterized in that: The construction and training process of the data prediction model is as follows: Building a data prediction model: ; In the formula, represents the constant value of the initial learner, is the target value of the nth training sample, N is the number of training samples, is the loss function, c is the constant when the loss function is minimized; For each sample The negative gradient is calculated using the following formula: ； In the formula, For the establishment of A regression tree for the sample The residual fit of For the The predicted value of a regression tree, is the learner constant; Based on data Fitting to get The leaf node area corresponding to the regression tree is ,in, , For the The number of leaf nodes in a regression tree, is the input quantity; for each leaf node Calculate the best fit value : ； The optimized data prediction model is expressed as: ; In the formula, is the indicator function.

8. The integrated river and lake ecological dredging monitoring method based on a hull according to claim 2 is characterized in that: The state prediction data at least includes: impurity removal state prediction data, adjustment state prediction data and pressing state prediction data; correspondingly, the abnormality prediction data at least includes: impurity removal abnormality prediction data, adjustment abnormality prediction data and pressing abnormality prediction data; The alarm model at least includes: a sludge impurity removal alarm model, a sludge regulation alarm model and a sludge pressing alarm model.

9. A hull-based integrated river and lake ecological dredging monitoring system, used to implement the river and lake integrated ecological dredging monitoring method as claimed in any one of claims 1 to 8, characterized in that: include: The hull is pre-arranged with a suction device, a sludge removal device, a sludge regulating tank, a sludge press and a mud cake conveying device according to the needs; It also includes: a plurality of weight sensors and vibration sensors, which are correspondingly installed in the sludge removal device; A plurality of first water level sensors, flow rate sensors and concentration sensors are correspondingly arranged at designated positions of the sludge regulating tank; A plurality of second water level sensors, pressure sensors and flow sensors are correspondingly arranged in the sludge press; And the first module is configured to obtain real-time data of the impurity removal status based on the sludge impurity removal device , the real-time data of the impurity removal status Inputting into the abnormal data analysis model to obtain the impurity removal abnormal data; creating a sludge impurity removal alarm model, determining the impurity removal abnormal points in the impurity removal link based on the impurity removal abnormal data and performing abnormal processing; The second module is configured to obtain the time series data of the regulation status of the sludge regulating tank , the adjustment state timing data Inputting into the abnormal data analysis model to obtain abnormal adjustment data; creating a sludge adjustment alarm model, determining the abnormal adjustment point in the adjustment link or impurity removal link based on the abnormal adjustment data and performing abnormal processing; The third module is configured to obtain the time series data of the squeezing state of the sludge press , the squeezing state time series data Input into the abnormal data analysis model to obtain the pressing abnormal data; create a sludge pressing alarm model, determine the pressing abnormal points in the pressing link, adjustment link or impurity removal link based on the pressing abnormal data and perform abnormal processing; The fourth module is configured to display and record the removal abnormal points, adjustment abnormal points and compression abnormal points in a visual form.

Images