Intelligent water sample detection workstation

Abstract

The invention discloses an intelligent water sample detection workstation which comprises an unmanned aerial vehicle and a ground station and has three functions of water sample collection and extraction, water sample detection and information data processing. The water sample collection and extraction part comprises a sampling disc carried by an unmanned aerial vehicle, a parking platform, a supply bin and a detection bin, and can realize automatic replacement of a sampling device and rapid extraction of a collected water sample; the water sample detection part comprises a peristaltic pump, a distilled water tank, a sampling and cleaning switching module, a spectrum water quality detection module, an electrode water quality detection module, a waste liquid pool and the like, rapid treatment and accurate analysis of a water sample can be realized, the intelligent level of water quality detection is greatly improved, and meanwhile, the efficiency and the accuracy of water quality detection are also improved; and the information data processing part comprises a host, a display screen, a communication module and a cooling fan. Meanwhile, a high-capacity lithium battery is arranged at the bottom of the work station, and the endurance requirement of the work station can be met.

Description

Technical Field

[0001] The invention relates to the field of water quality detection, in particular to an intelligent water sample detection workstation. Background Art

[0002] Water area monitoring systems usually adopt two solutions: fixed and mobile. With the increase of monitoring tasks and the expansion of monitoring areas, mobile solutions have become the focus of current research. Generally speaking, mobile monitoring platforms include unmanned ships, bionic fish, underwater robots, etc. Traditional unmanned ships are mostly based on dual propeller differential drive and can only travel in water. Since unmanned ships are relatively bulky, once they are used in one water area, if you want to change another water area for monitoring, there are shortcomings such as labor-intensive and dependent on lifting equipment. An amphibious unmanned ship (CN115742641A) that combines recovery function and underwater operation uses a specific recovery device to fix the unmanned ship to realize the land travel of the unmanned ship. In essence, the unmanned ship does not have amphibious functions. An amphibious water quality detection and sampling catamaran (CN115266239A) realizes amphibious operations by equipping underwater mobile components and land mobile components, but its water quality monitoring module is fixed and has not yet been able to realize multi-point and multi-water layer monitoring. It can be seen that the current unmanned boats can only monitor relatively fixed water layers, such as about 0.5m deep, when conducting water quality monitoring. They cannot go deep underwater to complete water quality monitoring at any location and any water layer. Existing solutions also include the use of drones for sampling and testing. For example, an amphibious sampling and testing drone (CN112414774A) achieves water operations by adding buoys to the drone bracket, but it only involves water sample collection and cannot go deep underwater. It can be seen that the existing unmanned boats and drone detection solutions have relatively simple functions. Summary of the invention

[0003] Purpose of the invention: The purpose of the present invention is to provide an intelligent water sample detection workstation that combines drones and ground stations, gives full play to their respective advantages, realizes water sample collection and rapid water quality detection, and the drone platform can collect samples in a large range and with high efficiency; the ground station can complete the automatic extraction of water samples collected by drones and the disassembly of sampling devices. The workstation integrates a spectral water quality detection module and an electrode water quality module, which can quickly and accurately detect multiple water quality indicators, such as dissolved oxygen, turbidity, ammonia nitrogen, etc., and provide comprehensive data support for water quality assessment.

[0004] Technical solution: The intelligent water sample detection workstation described in the present invention includes a drone and a ground station. The drone is equipped with a sampling tray at the bottom, and a number of sample bottles are arranged around the sampling tray for collecting water samples. A magnetic connector is provided on the top of the sampling tray. The ground station includes a shutdown loading and unloading unit, a sampling and cleaning switching unit, a spectral water quality detection module and an electrode water quality detection module. The shutdown loading and unloading unit removes the sampling tray containing the water sample by magnetic attraction and replaces it with an unused sampling tray. The sampling and cleaning switching unit takes water from the sampling tray containing the water sample by magnetic attraction and sends it to the electrode water quality detection module and the spectral water quality detection module for water quality detection. It is also used to clean the water intake pipe, the electrode water quality detection module and the spectral water quality detection module before taking water.

[0005] Furthermore, the sampling plate is adsorbed on the first power-off electromagnet of the telescopic rod at the bottom of the drone through a magnetic connector; when collecting water samples, the drone arrives at a designated location and hovers above the water surface, and the sampling plate is extended through the telescopic rod until it is submerged in the water. After a sample bottle is filled with water, the sampling plate is retracted through the telescopic rod, and the drone arrives at the next designated location to continue collecting water samples until all sample bottles on the sampling plate are filled with water.

[0006] Furthermore, the sample bottle includes a sealing cover connected by magnetic attraction. When the sample bottle collects water samples, the sealing cover is opened, and the sealing covers of other sample bottles are closed.

[0007] Furthermore, the parking and loading and unloading unit includes a parking platform located on the top of the ground station, which is the take-off and landing point of the UAV; position adjustment rods are arranged in the longitudinal and transverse directions of the parking platform. When the UAV lands on the parking platform, the position adjustment rods are used to push the UAV to a position directly above the replenishment bin; the replenishment bin contains unused sampling trays, and the sampling tray containing water samples falls onto the first unused sampling tray in the replenishment bin.

[0008] Furthermore, the shutdown loading and unloading unit also includes a detection bin and a sampling tray disassembly device, the sampling tray disassembly device includes a sampling rod, a nut seat is arranged on the sampling rod, and a second power-off type electromagnet is arranged on a section of the nut seat; the sampling tray containing the water sample is adsorbed from the replenishing bin by the second power-off type electromagnet, and the sampling tray containing the water sample is transferred from the replenishing bin to the detection bin through the nut seat, and the spectral water quality detection module and the electrode water quality detection module perform water quality detection on the water sample in the sampling tray containing the water sample in the detection bin.

[0009] Furthermore, a lifting machine is provided at the bottom of the replenishment bin and the inspection bin, which is used to raise the sampling plate to the bottom of the sampling rod to absorb the sampling plate, or to drop the sampling plate adsorbed on the sampling rod onto the lifting machine at the bottom of the inspection bin and then drop it to the bottom to complete the recovery of the sampling plate.

[0010] Furthermore, the sampling and cleaning switching unit includes a steering gear, a peristaltic pump and a sampling push device; the sampling push device includes a driving device and a water sampling needle;

[0011] After the workstation is started, the water sampling needle, spectral water quality detection module, electrode water quality detection module and corresponding pipelines are cleaned first. After sampling each sample bottle, the water sampling needle, spectral water quality detection module, electrode water quality detection module and corresponding pipelines are also cleaned. The cleaning process is as follows:

[0012] The driving device drives the water needle to align and push it into the second water suction port of the distilled water tank, sucks out the distilled water and sends it to the spectral water quality detection module and the electrode water quality detection module, and finally to the waste liquid pool, completing the cleaning of the water needle, the spectral water quality detection module, the electrode water quality detection module and the corresponding pipelines;

[0013] The distilled water after cleaning flows into the waste liquid pool.

[0014] Furthermore, when the sampling and cleaning switching unit is sampling, the sampling push device drives the water needle to align and push it into the first water suction port of the sample bottle through the driving device, and the water sample is sucked out by the peristaltic pump and sent to the spectral water quality detection module and the electrode water quality detection module, and then the water needle is withdrawn from the first water suction port;

[0015] After each sample bottle is sampled, the water sampling needle, the spectral water quality detection module, the electrode water quality detection module and the corresponding pipeline are cleaned, and then the next sample bottle is sampled until the sampling of the water sample in each sample bottle is completed;

[0016] The water samples detected in the spectral water quality detection module and the electrode water quality detection module flow into the waste liquid pool.

[0017] Furthermore, the spectral water quality detection module includes a full-spectrum light source, a spectrum analyzer and a cuvette, the slit of the spectrum analyzer is aligned with the full-spectrum light source, and the spectral intensity band of the full-spectrum light source through the cuvette is 254 to 1275 nm;

[0018] After the water sample enters the spectral water quality detection module, it flows through the cuvette. The spectrometer uses a machine learning algorithm to monitor the water quality of the water sample in the cuvette.

[0019] The spectral water quality detection module is provided with a black shell, a condenser is provided on the periphery of the full-spectrum light source, and a second cooling fan is provided on the back of the full-spectrum light source.

[0020] Furthermore, the electrode water quality detection module includes a pH sensor and a turbidity sensor. After the water sample enters the electrode water quality detection module, it contacts the detection positions of the pH sensor and the turbidity sensor to perform water quality monitoring.

[0021] Furthermore, after the spectral water quality detection module and the electrode water quality detection module complete the water quality detection, the detection results are displayed on the display screen, and the detection results are uploaded to the cloud through the mobile communication module.

[0022] Furthermore, the ground station is provided with a main compartment door for replenishing and recovering the sampling tray, tested water samples and cleaned wastewater; the ground station is also provided with a battery compartment and a battery compartment door.

[0023] Beneficial effects: Compared with the prior art, the advantages of the present invention are:

[0024] (1) The sampling device equipped in the present invention can automatically switch between sampling and cleaning states, so as to realize the rapid extraction of multiple water samples in the sampling plate and the automatic cleaning of the detection pipeline, eliminate the interference between the water samples, further improve the efficiency of water quality detection and reduce the interference between multiple water samples.

[0025] (2) The spectral water quality detection module and electrode water quality detection module equipped in the present invention can quickly and accurately detect multiple water quality data, providing comprehensive data support for water quality assessment.

[0026] (3) The parking platform designed in the present invention can automatically adjust the position of the returning drone, and the subsequent loading and unloading of the sampling tray can be completed without human intervention.

[0027] (4) The present invention is equipped with a sampling tray disassembly device, a supply chamber, and a detection chamber, which can autonomously realize the disassembly, loading, and recovery of the sampling tray carried by the UAV, further improving the efficiency of the workstation.

[0028] (5) The built-in micro-host of the present invention can realize the automatic control of the entire workstation and the real-time and accurate analysis and processing of the water quality data from the spectral water quality detection module and the electrode water quality detection module, and directly display the results on the display screen and synchronize the data results to the cloud.

[0029] (6) The present invention is provided with an independent battery compartment door and a main compartment door, which makes it more convenient for workers to replace batteries, replenish (recover) sampling plates, replenish distilled water, clean and replace waste liquid pools, and maintain cleaning devices, thereby further improving work efficiency.

[0030] (7) The water quality estimation method of the present invention has high accuracy and short modeling time, and can meet the requirements of rapid detection of outdoor water quality. Taking ammonia nitrogen as an example, the estimation method R 2 It reaches above 0.88, the RMSE is 0.0883, and the modeling time is only 0.1869s, which is much lower than 285.4631s and 11.3892s of 1DCNN and CAS-PLSR. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a front view of the workstation of the present invention;

[0032] Figure 2 It is a front perspective view of the workstation of the present invention;

[0033] Figure 3 A top view of the workstation of the present invention;

[0034] Figure 4 A top perspective view of a workstation of the present invention;

[0035] Figure 5 It is a schematic diagram of the sample bottle structure of the present invention;

[0036] Figure 6 A schematic diagram of the structure of the sampling and pushing device of the present invention;

[0037] Figure 7 It is a schematic diagram of the sampling state of the sampling and cleaning switching device of the present invention;

[0038] Figure 8 A schematic diagram of the cleaning state of the sampling and cleaning switching device of the present invention;

[0039] Fig. 9 A schematic diagram of the position of the distilled water tank of the present invention;

[0040] Fig.10 Schematic diagram of the structure of the invented distilled water tank;

[0041] Fig.11 A top perspective view of a spectral water quality detection module of the present invention;

[0042] Fig.12 It is a side perspective view of the spectral water quality detection module of the present invention;

[0043] Fig.13 It is a rear perspective view of the spectral water quality detection module of the present invention;

[0044] Fig.14 It is a front perspective view of the electrode detection module of the present invention;

[0045] Fig.15 A top perspective view of an electrode detection module of the present invention;

[0046] Fig.16 is a side perspective view of an electrode detection module of the present invention;

[0047] Fig.17 This is a schematic diagram of the drone of the present invention before the position is adjusted after landing on the parking platform;

[0048] Fig.18 This is a comparison chart of the ammonia nitrogen water quality estimation prediction value and the actual measured value of the present invention.

[0050] The parts numbers in the figure are as follows:

[0051] 1: UAV, 2: First power-off electromagnet, 3: Magnetic connector, 4: Sampling tray, 5: Main door, 6: Display screen, 7: Lithium battery, 8: Battery door, 9: Supplementary compartment, 10: Shock absorption area, 11: Second power-off electromagnet, 12: Connecting shaft, 13: Control processing compartment, 14: Main unit, 15: Display waterproof shell, 16: Sampling rod, 17: Sampling tray disassembly device, 18: Nut seat, 19: Threaded rod, 20: Servo motor, 21: Detection compartment, 22: Distilled water tank , 23: sampling push device, 24: sampling and cleaning switching device, 25: peristaltic pump, 26: spectral water quality detection module, 27: electrode water quality detection module, 28: latex water pipe, 29: water pipe connector, 30: waste liquid pool, 31: mobile communication module, 32: first cooling fan, 33: battery compartment, 34: lifting machine, 35: shock absorbing tray, 36: parking platform, 37: stepping motor, 38: longitudinal position adjustment rod, 39: lateral position adjustment rod, 40: servo motor slide;

[0052] 1-1: UAV landing gear, 4-1: sample bottle, 4-1-1: sample bottle silicone ring, 4-1-2: first water inlet, 4-1-3: catheter, 4-1-4: sealing cover, 22-1: water tank silicone ring, 22-2: second water inlet, 22-3: drainage tube, 23-1-1: No. 1 movable shaft, 23-1-2: No. 2 movable shaft, 23-2: servo steering gear, 23-3: sampling needle, 23-4: water needle fixing bracket, 23-5: linkage rod, 24-1: steering gear, 24-2: connecting bracket, 26-1: second cooling fan, 26-2: full spectrum light source, 26-2- 1: condenser, 26-3: first water outlet, 26-4: black frosted acrylic box, 26-5: spectrometer, 26-6: water inlet, 26-7: cuvette, 26-8: type-B, 27-1: RS485 interface, 27-2: module cover, 27-3: cover connection, 27-4: PH sensor, 27-5: second water outlet, 27-6-1: main water inlet, 27-6-2: auxiliary water inlet, 27-7: turbidity sensor, 27-8: connecting screw, 27-9: sealing rubber ring, 27-10: electrode water quality detection channel, 31-1: antenna. DETAILED DESCRIPTION

[0053] The technical solution of the present invention is further described below in conjunction with the accompanying drawings.

[0054] like Figures 1 to 4As shown, the intelligent water sample detection workstation combines drones and ground stations, giving full play to their respective advantages to realize the functions of water sample collection and rapid water quality detection. Through the drone platform, samples can be collected over a large range and with high efficiency. At the same time, the ground station can complete the automatic extraction of water samples collected by drones and the disassembly of the sampling device. The workstation integrates a spectral water quality detection module and an electrode water quality module, which can quickly and accurately detect multiple water quality indicators, such as dissolved oxygen, turbidity, ammonia nitrogen, etc., and provide comprehensive data support for water quality assessment.

[0055] The sampling tray 4 carried by the drone 1 is attracted by the telescopic rod 1-2 and the first power-off type electromagnet 2 on the drone 1 through the magnetic connector 3 and mounted on the drone 1. The magnetic connector 3 is located on the top of the sampling tray 4 to facilitate the replacement and loading and unloading of the sampling tray 4. The sample bottles 4-1 are distributed in a circular shape around the sampling tray 4. Figure 5 As shown, the sample bottle 4-1 includes a sample bottle silicone ring 4-1-1, a first water suction port 4-1-2, a conduit 4-1-3, and a sealing cover 4-1-4.

[0056] The ground work station includes a parking platform 36, a replenishment bin 9, a detection bin 21, a sampling tray disassembly device 17, a sampling push device 23, a sampling and cleaning switching device 24, a distilled water tank 22, a peristaltic pump 25, an electrode water quality detection module 27, a spectral water quality detection module 26, a waste liquid pool 30, a battery bin 33, a control and processing bin 13, a battery bin door 8, a main bin door 5, etc.

[0057] The parking platform 36 is at the top of the workstation and is the take-off and landing point of the drone 1. It includes a longitudinal position adjustment rod 38, a lateral position adjustment rod 39, and a servo motor slide 40. The parking platform 36 has two longitudinal position adjustment rods 38 and two lateral position adjustment rods 39 in the longitudinal and lateral directions, respectively. When the drone 1 returns to the parking platform 36, the servo motor slide 40 drives the position adjustment rod to push the drone 1 that has landed on the parking platform 36 to the top of the replenishment bin 9, so as to facilitate the subsequent replacement and removal of the sampling tray 4.

[0058] The sampling tray disassembly device 17 includes a servo motor 20, a threaded rod 19, a nut seat 18, a sampling rod 16, a connecting shaft 12, a stepper motor 37, and a second power-off type electromagnet 11 to move the sampling tray 4 that has completed sampling from the replenishment bin 9 to the detection bin 21 for detection. When the drone 1 lands on the parking platform 36 and completes the position calibration, the first power-off type electromagnet 2 on the drone 1 is energized, and the sampled sampling disk 4 will fall off to the first unused sampling disk 4 in the replenishment bin 9. At this time, the servo motor 20 will push the nut seat 18 to allow the sampling rod to be translated from the detection bin 21 to the supply bin 9. The front end of the sampling rod 16 is loaded with a stepper motor 37 and a second power-off type electromagnet 11 connected to the stepper motor 37. When the sampling rod 16 is directly above the sampled sampling disk 4 that has fallen off the drone 1 in the supply bin 9, the second power-off type electromagnet 11 at the front end of the sampling rod 16 is connected to the magnetic connector 3 on the sampling disk 16, and then the servo motor 20 translates the sampling rod 16 carrying the sampling disk 4 to the detection bin 21 through the nut seat 18 to carry out the next step of extracting water samples and testing.

[0059] The supply bin 9 is used to store unused sampling trays 4 and load unused sampling trays 4 to the drone 1. A lifting machine 34 is provided at the bottom of the supply bin 9, and a shock-absorbing tray 35 is provided at the top of the lifting machine 34. The unused sampling trays 4 are neatly stacked on the shock-absorbing tray 35 of the supply bin 9. When the sampled sampling trays 4 on the drone 1 are removed by the sampling tray removal device 17 and sent to the inspection bin 21 for inspection, the lifting machine 34 at the bottom of the supply bin 9 will extend upward by 40 cm, lift the first unused sampling tray 4 on the top to the first power-off electromagnet 2 of the drone 1, and the first power-off electromagnet 2 is connected to the magnetic connector 3 at the top of the sampling tray 4, and the lifting machine 34 will retract downward by 30 cm to complete the loading of the sampling tray 4.

[0060] like Figure 6 As shown, the sampling push device 23 realizes the pushing and retracting action of the water needle 23-3 by rotating the servo servo 23-2 in coordination with the No. 1 movable shaft 23-1-1, the linkage rod 23-5, the No. 2 movable shaft 23-1-2, and the water needle fixing bracket 23-4. The sampling push device 23 is fixed on the sampling and cleaning switching device 24 through the connecting bracket 24-2.

[0061] like Figure 7As shown, when the sampled sampling disk 4 is sent to the detection chamber 21, the sampling and cleaning switching device 24 is switched to the sampling state, that is, the direction of the water sampling needle 23-3 is adjusted by the servo 24-1 and aligned with the first water suction port 4-1-2 of the sample bottle 4-1 on the sampling disk 4. At this time, the sampling pushing device 23 pushes the water sampling needle 23-3 forward to the first water suction port 4-1-2 of the sample bottle 4-1, and is in close contact with the sample bottle silicone ring 4-1-1 to improve the airtightness and increase the sampling efficiency. At this time, the peristaltic pump 25 starts to work, and the water sample in the sample bottle 4-1 is sucked out and sent to the detection device through the latex water tube 28 for detection.

[0062] like Figure 8 As shown, when the detection device completes the detection, the sampling pushing device 23 removes the water collection needle 23-3 from the first water suction port 4-1-2 of the sample bottle 4-1, and the sampling cleaning switching device switches 24 to the cleaning state, that is, the direction of the sampling needle 23-3 is adjusted by the servo 24-1 and aligned with the water suction port of the distilled water tank 22. At this time, the sampling pushing device 23 pushes the water collection needle 23-3 forward into the second water suction port 22-2 of the distilled water tank, and is in close contact with the water tank silicone ring 22-1 to improve the airtightness and increase the extraction efficiency. At this time, the peristaltic pump 25 starts to work, and the distilled water in the distilled water tank 22 is sucked out through the drainage tube 22-3 and sent into the detection device through the latex water tube 28 and finally to the waste liquid pool 30 to achieve the cleaning effect of the water collection needle 23-3, the latex water tube 28, the peristaltic pump 25, the spectral water quality detection module 26, and the electrode water quality detection device 27. Fig. 9 and Fig.10 The position and structure of the distilled water tank 22 are shown.

[0063] After cleaning is completed, the sampling push device 23 pulls the water needle 23-3 out of the second water suction port 22-2 of the distilled water tank 22, and the sampling disk 4 rotates to the next sampling node of the sample bottle 4-1 through the stepper motor 37 above, and the sampling and cleaning switching device 24 switches to the sampling state, that is, the direction of the sampling needle 23-3 is adjusted through the servo 24-1 and aligned with the second water suction port 4-1-2 of the second sample bottle 4-1, and then the sampling steps of the first sample bottle 4-1 are repeated.

[0064] A lifting machine 34 is also provided at the bottom of the detection chamber 21, and a shock absorbing tray 35 is also provided at the top of the lifting machine 34. After the extraction and detection of all water samples are completed, the lifting machine 34 rises to the initial setting point, that is, 15 cm directly below the sampling tray 4, and the second power-off type electromagnet 11 in front of the sampling tray disassembly device 17 is energized to release the sampling tray 4, and the shock absorbing area 10 of the sampling tray 4 falls off onto the shock absorbing tray 35, and the lifting machine 34 descends 15 cm to complete the recovery of one sampling tray 4. The recovery steps of the sampling tray 4 are the same as above.

[0065] like Figures 11 to 13As shown, the spectral water quality detection module 26 includes a second cooling fan 26-1, a full-spectrum light source 26-2, a first water outlet 26-3, a black frosted acrylic box 26-4, a spectrum analyzer 26-5, a water inlet 26-6, a cuvette 26-7, and a water pipe connector 29. The housing of the spectral water quality detection module 26 is composed of a black frosted acrylic plate 26-4, which can effectively reduce the interference of natural light on the spectrum analyzer 26-5. The slit of the spectrometer 26-5 is aligned with the full-spectrum light source 26-2 to ensure the aggregation of the light source. At the same time, the full-spectrum light source 26-2 is provided with a condenser 26-2-1 on the periphery and a second cooling fan 26-1 on the back to ensure the pertinence of the light source and reduce the detection error caused by the heat of the light source. A cuvette 26-7 is placed between the two. When the extracted water sample is transported to the water pipe connection 29 of the spectral water quality detection module 26 through the peristaltic pump 25 and the latex water tube 28, the water sample flows into the module and flows through the cuvette 26-7. At this time, the spectrometer 26-5 will analyze the water sample and transmit the obtained relevant water quality data to the host 14 through the type-B26-8 interface on the spectrometer 26-5, and then flows out of the spectral water quality detection module 26 through the latex water tube 28. The spectrometer 26-5 collects the full spectrum light source 26-2 of the tungsten lamp, and the spectral intensity band of the quartz flow cuvette 26-7 is 254~1275nm, with a total of 1920 spectral data. Then the machine learning algorithm constructs a water quality estimation model for water quality testing.

[0066] like Figures 14 to 16 As shown, the electrode water quality detection module 27 includes an RS485 interface 27-1, a module upper cover 27-2, an upper cover connection 27-3, a PH sensor 27-4, a second water outlet 27-5, a main water inlet 27-6-1, a secondary water inlet 27-6-2, a turbidity sensor 27-7, a connecting screw 27-8, a sealing rubber ring 27-9, and a water pipe connector 29. The main water inlet 27-6-1 and the secondary water inlet 27-6-2 can be connected to different water inlets according to the size of different sensors. The electrode water quality detection module 27 is an open and detachable design, and users can also replace the sensors required to measure data according to their needs. After the water sample is detected by the spectral water quality detection module 26, it flows into the water inlet through the latex water tube 28. When the water sample flows through the electrode water quality detection channel 27-10, it contacts the sensor detection position. The corresponding sensor detects the water sample and connects the data to the host through the RS485 interface 27-1. The water sample then flows out from the second water outlet 27-5.

[0067] The waste liquid pool 30 is connected to the second water outlet 27-5 of the electrode detection module 27 through the latex water pipe 28. When the water sample flows out from the second water outlet 27-5, all water quality data detection is completed and the water sample flows into the waste liquid pool 30 through the latex water pipe 28. Similarly, the distilled water after cleaning the water sampling needle 23-3, the latex water pipe 28, the peristaltic pump 25, the spectral water quality detection module 26, and the electrode water quality detection module 27 will also flow into the waste liquid pool 30.

[0068] The control processing chamber 13 includes a host 14, a display screen 6, a display screen waterproof shell 15, a first cooling fan 32, and a mobile communication module 31 to realize the control of the system and the processing and transmission of the collected data. The host 14 controls the normal operation of the servo motor slide 40, the sampling plate disassembly device 17, the stepper motor 37, the servo motor 20, the second power-off type electromagnet 11, the lifting machine 34, the steering gear 24-1, the sampling push device 23, and the peristaltic pump 25. The host 14 will also process and analyze the water quality data transmitted by the spectral water quality detection module 26 and the electrode water quality detection module 27. The processed results will be synchronized to the display screen 6, and the data will be uploaded to the cloud through the mobile communication module 31. The control processing chamber 13 has a built-in first cooling fan 32 and a heat dissipation hole to dissipate the heat of the host 14 and the display screen 6 to obtain the best processing performance.

[0069] The battery compartment 33 contains a large-capacity lithium battery 7 to power the intelligent water sample detection workstation. The battery compartment 33 is provided with a battery compartment door 8 for the convenience of users in replacing and replenishing batteries. At the same time, the battery compartment door is provided with heat dissipation holes, which can effectively reduce the temperature of the lithium battery 7 and extend the service life of the lithium battery 7.

[0070] The main door 5 is located on the right side of the display screen 6 and the heat dissipation hole, and is used to replenish the sampling tray 4 and the distilled water tank 22 and to recycle and process the used sampling tray 4 and the waste liquid pool 30. It is also convenient for the staff to perform daily maintenance and cleaning of the equipment.

[0071] The working principle of the workstation of the present invention includes the following steps:

[0072] S1. Preparation stage: The staff unfolds the antenna 31 - 1 of the intelligent water sample detection workstation (hereinafter referred to as the workstation), opens the battery compartment door 8 and the main compartment door 5 , installs the lithium battery 7 and 6 unused sampling trays 4 , a full distilled water tank 22 , and a waste liquid pool 30 .

[0073] S2. After turning on the power of the workstation, the workstation will perform a self-test for 3 minutes, mainly including checking whether the servo motor slide 40, the servo motor 20, the stepper motor 37, the second power-off type electromagnet 11, the sampling and cleaning switching device 24, the sampling pushing device 23, the peristaltic pump 25, the two lifting and lowering platforms 34, and the first cooling fan 32 are operating normally, and obtaining the current power of the workstation, the number of unused sampling trays 4, the number of recovered sampling trays 4, the distilled water level, the water level of the waste liquid pool 30 and displaying the status on the display screen 6.

[0074] When a component is found to be abnormal, the battery power is low, the number of unused sampling trays 4 is insufficient, the number of recovered sampling trays 4 is full, the distilled water is insufficient, or the water level in the waste liquid pool 30 is too high, the display screen 6 will push a corresponding warning message, and the workstation will not be able to work normally.

[0075] After the workstation self-inspection is completed and there is no error, the sampling and cleaning switching device 24 switches to the cleaning state for self-cleaning, that is, the direction of the sampling needle 23-3 is adjusted by the servo 24-1 and aligned with the second water suction port 22-2 of the distilled water tank 22. At this time, the sampling pushing device 23 pushes the water needle 23-3 into the second water suction port 22-2, and is in close contact with the water tank silicone ring 22-1 to improve the airtightness and increase the extraction efficiency. The peristaltic pump 25 draws the distilled water out of the distilled water tank 22 and sends it into the detection device through the latex water pipe 25 and finally to the waste liquid pool 30, so as to achieve the cleaning effect and discharge the water quality interference items in the detection device caused by long-term non-use.

[0076] During the cleaning, the spectral water quality detection module 26 and the electrode water quality detection module 27 will be initialized and calibrated to achieve more accurate subsequent water sample detection.

[0077] S3. After the workstation completes the self-check and self-cleaning operations, the display screen 6 will display "The workstation can work normally". At this time, the staff will put the drone 1 on the parking platform 36, turn it on and initialize the return point of the drone 1. The unused sampling trays 4 are neatly stacked in the replenishment bin 9 in sequence. The lifting machine 34 at the bottom of the replenishment bin 9 will extend upward by 45 cm, lift the first unused sampling tray 4 on the top to the first power-off electromagnet 2, and the first power-off electromagnet 2 will be connected to the magnetic connector 3 on the top of the sampling tray 4. The lifting machine 34 will retract 30 cm downward, completing the loading of the sampling tray 4 and reducing the number of unused sampling trays by 1.

[0078] S4. The staff controls the drone 1 to collect water samples. When the drone arrives at the designated water sample collection area, the drone hovers 30cm above the water surface, and then the telescopic rod 1-2 on the drone connects the first power-off type electromagnet 2, the magnetic connector 3 and the sampling plate 4. The telescopic rod 1-2 extends downward until the sampling plate 4 is submerged in the water, and the sealing cover 4-1-4 of the first sample bottle 4-1 is opened. At this time, the sealing covers of other sample bottles 4-1 are in a closed state. After the first sample bottle 4-1 is filled with water through the conduit 4-1-3, the telescopic rod 1-2 is retracted, the sampling plate 4 is retracted to the water surface, and the sealing cover 4-1-4 is closed. The drone completes a water sample collection; after the drone flies to the next location, the sampling process of the sample bottle is the same as above, and a bottle of water sample is collected at each location.

[0079] S5. After the drone 1 carrying the sampling tray 4 completes the sampling task of all the sample bottles 4-1, the drone 1 automatically returns to the destination according to the GPS positioning and lands on the parking platform 36 of the workstation. Due to the allowable error of GPS positioning, it may not be able to land exactly above the replenishment bin 9, that is, the following situation occurs: Fig.17 After the workstation receives the information that the drone 1 has landed on the parking platform 36, the host 14 controls the servo motor slide 40 in four directions to drive the longitudinal position adjustment rod 38 and the lateral position adjustment rod 39 to slowly surround the top of the replenishment bin 9 in the middle of the parking platform 36. With the four position adjustment rods in different directions slowly pushing the drone landing gear 1-1, the drone 1 moves accurately to the top of the replenishment bin 9, completing the initial adjustment before sampling and waiting for the next operation of the workstation.

[0080] S6. After the workstation completes the adjustment of the position of the drone 1 before sampling, the workstation sends a "release sampling tray" command to the drone 1 through the host 14. When the drone 1 receives the command, it will energize the first power-off electromagnet 2. At this time, the magnetic connector 3 above the sampling tray 4 loses its magnetic attraction and automatically falls off to the top unused sampling tray 4 in the replenishment bin 9 under the action of gravity.

[0081] S7. After the workstation sends the "release sampling disk" command 8s later, the host 14 controls to push the nut seat 18 to allow the sampling rod 16 to move from the detection chamber 21 to the supply chamber 9. The front end of the sampling rod 16 is loaded with a stepper motor 37 and a second power-off electromagnet 11 connected to the stepper motor 37. When the sampling rod 16 is directly above the sampled sampling disk 4 that has fallen off the drone 1 in the replenishment chamber 21, the power-off electromagnet 11 at the front end of the sampling rod 16 is connected to the magnetic connector 3 on the sampling disk 4 through magnetic attraction, and then the servo motor 20 moves the nut seat 18 to translate the sampling rod 16 mounted with the sampling disk 4 into the detection chamber 21 for the next step of extracting water samples and testing.

[0082] S8. While the sampling tray 4 is being sent into the detection chamber 21, the staff replaces the battery of the drone 1, re-powers it, and initializes the drone's return point; the lifting machine 34 at the bottom of the supplementary chamber 9 will extend upward by 45 cm, and lift the first unused sampling tray 4 at the top to the first power-off type electromagnet 2. The first power-off type electromagnet 2 is connected to the magnetic connector 3 at the top of the sampling tray 4, and the lifting machine 34 shrinks downward by 30 cm, completing one loading of the sampling tray 4 and reducing the number of unused sampling trays 4 by 1. The loading steps of the subsequent sampling trays 4 are the same as above. When the number of unused sampling trays 4 is 0, the workstation will automatically lock and cannot continue to test. The staff needs to open the main compartment door 5 to take out the recovered sampling trays 4 and replenish the corresponding number of unused sampling trays 4. After the self-test of the workstation is performed through the button below the screen 6, the detection work can continue. At this time, the staff can control the drone 1 again to continue collecting water samples.

[0083] S9, when the sampled sampling disk 4 is sent to the detection chamber 21, the sampling and cleaning switching device 24 is switched to the sampling state, and the water sample in the sample bottle 4-1 is sucked out and sent to the water pipe connection 29 of the spectral water quality detection module 26 through the latex water pipe 28, and then the water sample flows into the module, flows through the cuvette 26-7, and then the spectrum analyzer 26-5 will analyze the water sample, and transmit the acquired relevant water quality data to the host 14 through the type-B interface 26-8 on the spectrum analyzer 26-5, and the host 14 will process and analyze the water quality through the spectrum information, and finally obtain the water quality data result and store it in advance. At this time, the water sample flows out of the spectral water quality detection module 26 through the latex water pipe 28 and flows into the water inlet of the electrode water quality detection module 27. When the water sample flows through the module, it contacts the sensor detection position, and the corresponding sensor detects the water sample, and transmits the data to the host 14 through the RS485 interface 27-1 connecting cable. The host 14 processes and analyzes the acquired data, compares it with the water quality data pre-stored in the spectral water quality detection module 26 to eliminate erroneous data and improve the accuracy of the data, and finally displays all the processed water quality data on the display screen 6 and uploads the data to the cloud platform through the mobile communication module 31.

[0084] S10, the water sample flows out of the electrode water quality detection module 27, and finally flows into the waste liquid pool 30 through the latex water pipe 28. When the water level of the waste liquid pool 30 is higher than the preset value, the workstation will be automatically locked and cannot be tested. The staff needs to open the main compartment door 5, replace the waste liquid pool 30 or pour out the waste liquid and clean it, and then replace the distilled water tank 22 or add distilled water. After the self-check of the workstation is performed by pressing the button below the screen 6, the detection work can be continued.

[0085] S11. After the test is completed, the sampling push device 23 removes the water needle 23-3 from the first water suction port 4-1-2 of the sample bottle 4-1, the sampling and cleaning switching device 24 switches to the cleaning state, and the peristaltic pump 25 starts to work for 15 seconds, sucking the distilled water in the distilled water tank 22 and sending it to the detection device through the latex water tube 28 and finally to the waste liquid pool 30 to achieve the cleaning effect of the water needle 23-3, the latex water tube 28, the peristaltic pump 25, the spectral water quality detection module 26, and the electrode water quality detection module 27. When the distilled water level is lower than the minimum water level, the workstation will automatically lock and cannot be tested. The staff needs to open the main compartment door 5, replace the distilled water tank 22 or replenish distilled water, and then replace or pour out the waste liquid in the waste liquid pool 30 and clean it. After the self-check of the workstation is performed by the button below the screen 6, the detection work can continue.

[0086] S12. After cleaning is completed, the sampling push device 23 removes the water needle 23-3 from the second water suction port 22-2 of the distilled water tank 22, and the sampling disk 4 is rotated to the next sampling node of the sample bottle 4-1 through the stepper motor 37 above. The sampling and cleaning switching device 24 is switched to the sampling state, that is, the direction of the sampling needle 23-3 is adjusted through the servo 24-1 and aligned with the water suction port 4-1-2 of the second sample bottle 4-1, and then the sampling steps of the first sample bottle 4-1 are repeated.

[0087] S13, when the sampling device completes the extraction and detection of all water samples, the lifting machine 34 at the bottom of the detection chamber 21 rises to the initial setting point (15 cm directly below the sampling plate), and the second power-off type electromagnet 11 in front of the sampling plate disassembly device 17 is energized to release the sampling plate 4, and the sampling plate 3 falls off onto the lifting machine 34, and the lifting machine 34 drops 15 cm, completing the recovery of one sampling plate 4. At this time, the number of recovered sampling plates 4 on the screen 6 is increased by 1, and the subsequent recovery steps of the sampling plates 4 are the same as above. When the number of recovered sampling plates 4 reaches 6, the workstation will automatically lock and cannot continue to detect. The staff needs to open the main chamber door 5 to take out the recovered sampling plates 4 and replenish the corresponding number of unused sampling plates 4. After that, the self-check of the workstation can be continued through the button below the screen 6.

[0088] S14. After completing all water sample detection tasks, the staff clicks "End Task" on the screen, and the workstation will perform a self-cleaning operation (the steps are the same as the "self-cleaning" operation described in S2). After completing the "self-cleaning" operation, the staff turns off the power, opens the battery compartment door 8 and the main compartment door 5, takes out the lithium battery 7 and all unused sampling trays 4, recycles the sampling tray 4, the distilled water tank 22 and the waste liquid tank 30, and then folds the antenna 33-1.

[0089] The comparison diagram of the estimated predicted value and the measured value of ammonia nitrogen water quality in this embodiment is as follows Fig.18As shown, in this embodiment, the spectrum analyzer (26-5) collects the spectral intensity band of the tungsten lamp through the quartz flow cuvette (26-7) in the range of 254 to 1275 nm, with a total of 1920 spectral data. The data are preprocessed by a standardized method, and then the principal component analysis method is used to screen the characteristic spectra to 400. For ammonia nitrogen in different bands, the detection accuracy obtained by using a variety of spectral preprocessing methods and machine learning models is shown in Tables 1 to 3. It can be seen that the detection accuracy of the present invention is relatively high.

[0090] Table 1 Comparison of machine learning model accuracy under different spectral preprocessing methods for full-band ammonia nitrogen

[0091]

[0092] Note 1D-CNN: one-dimensional convolutional neural network; CAS-PLSR: a fusion algorithm of lock-free non-blocking algorithm and partial least squares method; PLSR: partial least squares method; RF: random forest algorithm; RF-L: a fusion algorithm of random forest algorithm and LASSO algorithm; SVM-ML: a fusion algorithm of support vector machine (SVM) and multi-layer perceptron (MLPRegressor); SVR: support vector machine regression algorithm; the same below.

[0093] Table 2 Comparison of machine learning model accuracy under different spectral preprocessing methods for ammonia nitrogen in 1000 characteristic bands

[0094]

[0095] Table 3 Comparison of machine learning model accuracy under different spectral preprocessing methods for ammonia nitrogen in 400 characteristic bands

[0096]

Claims

1. An intelligent water sample detection workstation, characterized in that: The invention comprises an unmanned aerial vehicle (1) and a ground station. The bottom of the unmanned aerial vehicle (1) is equipped with a sampling tray (4). A plurality of sample bottles (4-1) are arranged around the sampling tray (4) for collecting water samples. The top of the sampling tray (4) is provided with a magnetic connector (3). The ground station comprises a shutdown loading and unloading unit, a sampling cleaning switching unit, a spectral water quality detection module (26) and an electrode water quality detection module (27). The shutdown loading and unloading unit removes the sampling tray (4) containing the water sample by magnetic attraction and replaces it with an unused sampling tray (4). The sampling cleaning switching unit takes water from the sampling tray (4) containing the water sample by magnetic attraction and sends it to the electrode water quality detection module (27) and the spectral water quality detection module (26) for water quality detection. The unit is also used to clean the water taking pipe, the electrode water quality detection module (27) and the spectral water quality detection module (26) before taking water.

2. The intelligent water sample detection workstation according to claim 1, characterized in that: The sampling plate (4) is adsorbed on the first power-off type electromagnet (2) of the telescopic rod (1-2) at the bottom of the drone (1) via a magnetic connector (3); when collecting water samples, the drone (1) arrives at a designated location and hovers above the water surface, and the sampling plate (4) is extended through the telescopic rod (1-2) until it is submerged in the water surface. After a sample bottle (4-1) is filled with water, the sampling plate (4) is retracted through the telescopic rod (1-2), and the drone (1) arrives at the next designated location to continue collecting water samples until all sample bottles (4-1) of the sampling plate (4) are completely filled with water.

3. The intelligent water sample detection workstation according to claim 2 is characterized in that: The sample bottle (4-1) comprises a sealing cover (4-1-4) connected by magnetic attraction. When the sample bottle (4-1) collects water samples, its sealing cover (4-1-4) is opened, while the sealing covers (4-1-4) of the other sample bottles (4-1) are closed.

4. The intelligent water sample detection workstation according to claim 1, characterized in that: The parking and loading and unloading unit comprises a parking platform (36) located at the top of the ground station, which is the take-off and landing point of the UAV (1); the parking platform (36) is provided with position adjustment rods in the longitudinal and transverse directions, and when the UAV (1) falls on the parking platform (36), the position adjustment rods are used to push the UAV (1) to be located directly above the replenishment bin (9); the replenishment bin (9) contains unused sampling trays (4), and the sampling tray (4) containing water samples falls onto the first unused sampling tray (4) in the replenishment bin (9).

5. The intelligent water sample detection workstation according to claim 4, characterized in that: The shutdown loading and unloading unit further comprises a detection chamber (21) and a sampling tray disassembly device (17), wherein the sampling tray disassembly device (17) comprises a sampling rod (16), a nut seat (18) is arranged on the sampling rod (16), and a second power-off type electromagnet (11) is arranged on a section of the nut seat (18); the sampling tray (4) containing the water sample is adsorbed from the replenishment chamber (9) by the second power-off type electromagnet (11), and the sampling tray (4) containing the water sample is transferred from the replenishment chamber (9) to the detection chamber (21) by the nut seat (18); and the spectral water quality detection module (26) and the electrode water quality detection module (27) perform water quality detection on the water sample in the sampling tray (4) containing the water sample in the detection chamber (21).

6. The intelligent water sample detection workstation according to claim 5, characterized in that: The bottom of the replenishing chamber (9) and the detection chamber (21) are both provided with a lifting machine (34) for lifting the sampling plate (4) to the bottom of the sampling rod (16) to absorb and load the sampling plate (4), or for dropping the sampling plate (4) adsorbed on the sampling rod (16) onto the lifting machine (34) at the bottom of the detection chamber (21) and then dropping it to the bottom to complete the recovery of the sampling plate (4).

7. The intelligent water sample detection workstation according to claim 1, characterized in that: The sampling and cleaning switching unit comprises a steering gear (24-1), a peristaltic pump (25) and a sampling pushing device (23); the sampling pushing device (23) comprises a driving device and a water sampling needle (23-3); After the workstation is started, the water sampling needle (23-3), the spectral water quality detection module (26), the electrode water quality detection module (27) and the corresponding pipelines are cleaned first. After sampling each sample bottle (4-1), the water sampling needle (23-3), the spectral water quality detection module (26), the electrode water quality detection module (27) and the corresponding pipelines are also cleaned. The cleaning process is as follows: The driving device drives the water collection needle (23-3) to align with and push it into the second water suction port (22-2) of the distilled water tank, sucks out the distilled water and sends it to the spectral water quality detection module (26) and the electrode water quality detection module (27), and finally to the waste liquid pool, thereby completing the cleaning of the water collection needle (23-3), the spectral water quality detection module (26), the electrode water quality detection module (27) and the corresponding pipelines; The distilled water after cleaning flows into the waste liquid tank (30).

8. The intelligent water sample detection workstation according to claim 7, characterized in that: When the sampling and cleaning switching unit is sampling, the sampling pushing device (23) drives the water sampling needle (23-3) through the driving device to align with and push it into the first water suction port (4-1-2) of the sample bottle (4-1), and the water sample is sucked out by the peristaltic pump (25) and sent to the spectral water quality detection module (26) and the electrode water quality detection module (27), and then the water sampling needle (23-3) is withdrawn from the first water suction port (4-1-2); After each sampling of a sample bottle (4-1) is completed, the water sampling needle (23-3), the spectral water quality detection module (26), the electrode water quality detection module (27) and the corresponding pipelines are cleaned, and then the next sample bottle (4-1) is sampled, until the sampling of the water sample in each sample bottle (4-1) is completed; The water sample that has completed the detection in the spectral water quality detection module (26) and the electrode water quality detection module (27) flows into the waste liquid pool (30).

9. The intelligent water sample detection workstation according to claim 1, characterized in that: The spectral water quality detection module (26) comprises a full-spectrum light source (26-2), a spectrum analyzer (26-5) and a cuvette (26-7), wherein the slit of the spectrum analyzer (26-5) is aligned with the full-spectrum light source (26-2), and the spectral intensity band of the full-spectrum light source (26-2) passing through the cuvette (26-7) is 254 to 1275 nm; After the water sample enters the spectral water quality detection module (26), it flows through the cuvette (26-7), and the spectral analyzer (26-5) uses a machine learning algorithm to monitor the water quality of the water sample in the cuvette (26-7); The spectral water quality detection module (26) is provided with a black shell, a condenser (26-2-1) is provided on the periphery of the full-spectrum light source (26-2), and a second cooling fan (26-1) is provided on the back of the full-spectrum light source (26-2).

10. The intelligent water sample detection workstation according to claim 1, characterized in that: The electrode water quality detection module (27) comprises a pH sensor (27-4) and a turbidity sensor (27-7). After the water sample enters the electrode water quality detection module (27), it contacts the detection positions of the pH sensor (27-4) and the turbidity sensor (27-7) to perform water quality monitoring.

11. The intelligent water sample detection workstation according to claim 1, characterized in that: After the spectral water quality detection module (26) and the electrode water quality detection module (27) complete the water quality detection, the detection results are displayed on the display screen (6), and the detection results are uploaded to the cloud through the mobile communication module (31).

12. The intelligent water sample detection workstation according to claim 1, characterized in that: The ground station is provided with a main compartment door (5) for replenishing and recovering the sampling tray (4), tested water samples and cleaned wastewater; the ground station is also provided with a battery compartment (33) and a battery compartment door (8).

Images