Mobile water quality detection system

Abstract

The invention discloses a water quality mobile detection system, belongs to the technical field of water quality detection, and solves the problems of sparse sailing data, partial cognition of a diffusion model, mismatching of emergency resources and threat of ecological and drinking water safety due to secret drainage and hydrological dynamics. Comprising a master control and calculation unit which is used for overall planning and executing intelligent decision making and control of the whole process of environment calibration, strategy generation, data inversion, decision making scheduling and model learning; the active disturbance generation unit is used for receiving and executing an instruction from the main control unit and applying customized quantum optics, a specific chemical tracer agent, a structured turbulent flow field and other multi-mode physical and chemical disturbances to the water body; and the multi-mode signal acquisition and processing unit. According to the method, controlled physical or chemical disturbance is applied to a water body, multi-dimensional response signals of the water body are collected and analyzed in real time, the position and the diffusion trend of a pollution source are dynamically inverted, and closed-loop learning is carried out after a task is completed so as to optimize subsequent performance.

Description

Technical Field

[0001] This invention relates to the field of water quality testing technology, and in particular to a mobile water quality testing system. Background Technology

[0002] Mobile water quality monitoring enables rapid, real-time water quality analysis by moving equipment to the water source or along the water body. Its core features are mobile equipment and real-time data, overcoming the spatial and temporal limitations of fixed laboratories and online monitoring stations, and dynamically monitoring water quality changes. It is mainly divided into four categories: portable on-site testing, vehicle-mounted and ship-mounted mobile laboratories, unmanned intelligent monitoring platforms, and miniaturized fixed and semi-fixed monitoring stations. These are carried in various formats, such as suitcases, modified vehicles, unmanned boats, and micro-stations, to suit different scenarios. Portable equipment is flexible and lightweight, mobile laboratories offer strong testing capabilities, unmanned platforms can move autonomously, and micro-monitoring stations can form a monitoring network. With its mobility and real-time advantages, this technology can be used for routine inspections and emergency responses to sudden pollution events, and is developing towards intelligence, integration, and networking in the future.

[0003] When tracing the source of pollution in the downstream river of an industrial park, the intermittent nature of pollution discharge and the dynamic nature of hydrological conditions during mobile water quality monitoring make it difficult to capture the full picture of a pollution event through a single instantaneous "snapshot". This sparsity of data in the spatiotemporal dimension makes it impossible to transform its authenticity into effective representativeness.

[0004] When these sparse and potentially lagging data are input into diffusion models that rely on fixed parameters, the models may develop severe cognitive biases due to incorrect location inputs and distorted hydrological conditions, resulting in a distorted "pollution map".

[0005] The direct consequence is that the resource deployments made by emergency command departments based on this erroneous map, such as the deployment of fences and the addition of treatment agents, are completely deviated from the actual location and diffusion path of the pollution plume. This not only wastes valuable emergency resources and delays critical disposal opportunities, but may also lead to the continued spread of uninterrupted pollutants downstream, posing a substantial long-term threat to drinking water safety or sensitive ecosystems.

[0006] Therefore, a mobile water quality monitoring system is proposed to solve or alleviate the above problems. Summary of the Invention

[0007] The purpose of this invention is to address the shortcomings of existing technologies by proposing a mobile water quality detection system.

[0008] To achieve the above objectives, the present invention adopts the following technical solution: A mobile water quality monitoring system, including The main control and computing unit is used to coordinate and execute intelligent decision-making and control throughout the entire process of environment calibration, strategy generation, data inversion, decision scheduling and model learning; The active disturbance generation unit is used to receive and execute instructions from the main control unit to apply customized quantum optics, specific chemical tracers and structured turbulent fields and other multimodal physicochemical disturbances to the water body. The multimodal signal acquisition and processing unit is used to synchronously acquire multidimensional response signals such as quantum state correlation, flow field structure distortion and spectral fingerprint of water body in response to disturbance events, and to perform real-time preprocessing and feature extraction. The power supply and energy management unit is used to provide power to each unit and integrate energy recovery; The communication and clock synchronization unit is responsible for time synchronization and data communication to ensure that instructions and data between distributed units are consistent in time and space.

[0009] Preferably, the data bus interface of the main control and computing unit is electrically connected to the data output interface of the multimodal signal acquisition and processing unit; the control signal port of the main control and computing unit is electrically connected to the controlled end of the active disturbance generation unit; the data interface of the communication and clock synchronization unit is electrically connected to the communication interface of the main control and computing unit; and the first, second, and third regulated output terminals of the power supply and energy management unit are electrically connected to the power input terminals of the main control and computing unit, the active disturbance generation unit, and the multimodal signal acquisition and processing unit, respectively.

[0010] Preferably, the active perturbation generating unit includes a quantum coherent excitation module, which includes a laser diode driver chip, a temperature control chip, a nonlinear optical crystal, a laser diode, a thermistor, and a thin-film heater. The enable pin of the laser diode driver chip is electrically connected to the general-purpose input / output port of the main control and computing unit to receive an enable control signal. The feedback pin of the laser diode driver chip is grounded through a first sampling resistor. The switch output pin of the laser diode driver chip is connected to the anode of the laser diode. The cathode of the laser diode is grounded. The nonlinear optical crystal is used to generate entangled photon pairs under the action of the pump light emitted by the laser diode. A thermistor is mounted on the surface of a nonlinear optical crystal. The two ends of the thermistor are connected to the positive and negative input pins of the temperature sensor signal of the temperature control chip, respectively. A thin-film heater is attached to the surface of the nonlinear optical crystal. One end of the thin-film heater is connected to the power supply, and the other end is connected to the heater drive pin of the temperature control chip. The communication interface of the temperature control chip is electrically connected to the serial communication port of the main control and computing unit to receive the temperature setpoint and return the temperature status. The synchronization signal output pin of the laser diode driver chip is electrically connected to the start channel input pin of the time-to-digital converter in the multimodal signal acquisition and processing unit to provide the excitation event start timestamp.

[0011] Preferably, the active perturbation generating unit further includes a controllable chemical perturbation module. The controllable chemical perturbation module includes a first motor drive chip, a high-voltage pulse generator chip, a micro-injection pump, and a microcavity electroporation release device. The logic input pin of the first motor drive chip is electrically connected to the pulse width modulation output port and direction control pin of the main control and computing unit to receive motion control commands. The power input pin of the first motor drive chip is electrically connected to the motor drive power output terminal provided by the power supply and energy management unit. The bridge output pin of the first motor drive chip is connected to the two winding terminals of the DC motor in the micro-injection pump to drive its piston to reciprocate to inject or extract liquid tracer. The external trigger pin of the high-voltage pulse generator chip is connected to the main control and computing unit... The general-purpose input / output port is electrically connected to receive a discharge trigger signal. The high-voltage power supply pin of the high-voltage pulse generator chip is connected to a high-voltage DC power supply through a current-limiting resistor. The ground pin of the high-voltage pulse generator chip is grounded. The high-voltage output pin of the high-voltage pulse generator chip is connected to the high-voltage electrode of the microcavity electroporation release device. The ground electrode of the microcavity electroporation release device is grounded. The internal cavity of the microcavity electroporation release device is used to contain capsules encapsulated with engineered microorganisms and generates a strong electric field when receiving a high-voltage pulse to break down the capsule membrane and achieve controlled release. The fluid outlet of the micro-injection pump is used to connect to an injection nozzle located in the water body through a pipeline. The fault status output pin of the first motor drive chip is electrically connected to the interrupt input pin of the main control and computing unit for feedback of the drive status.

[0012] Preferably, the active disturbance generation unit further includes a synthetic turbulence field module, which includes a three-phase full-bridge driver chip, first to sixth power MOSFETs, first to third bootstrap diodes, first to third bootstrap capacitors, a three-phase magnetohydrodynamic thruster, a current detection chip, and a thermistor. The three high-side logic input pins and three low-side logic input pins of the three-phase full-bridge driver chip are electrically connected to the pulse width modulation signal output port of the main control and computing unit to receive six independent control waveforms. The three high-side gate drive pins of the three-phase full-bridge driver chip are connected to the gates of the first to third power MOSFETs through first to third gate resistors. The three low-side gate drive pins of the three-phase full-bridge driver chip are connected to the gates of the fourth to sixth power MOSFETs through fourth to sixth gate resistors. The anodes of the first to third bootstrap diodes are connected to the driving power supply, and their cathodes are connected to the first to third high-side floating power supply pins of the three-phase full-bridge driver chip. One end of each of the first to third bootstrap capacitors is connected to the first to third high-side floating power supply pins of the three-phase full-bridge driver chip. The other ends of the first to third bootstrap capacitors are respectively connected to the first to third high-side reference voltage pins of the three-phase full-bridge driver chip. The first to third high-side reference voltage pins are respectively electrically connected to the sources of the fourth to sixth power MOSFETs. The drains of the first to third power MOSFETs are commonly connected to the high-voltage DC bus. The sources of the first to third power MOSFETs serve as the three-phase output terminals. The drains of the fourth to sixth power MOSFETs are respectively electrically connected to the three-phase output terminals. The sources of the fourth to sixth power MOSFETs are commonly connected to the current detection pin of the current detection chip and grounded. The three-phase output terminals are respectively connected to the three coil input terminals of the three-phase magnetohydrodynamic thruster. The voltage output pin of the current detection chip is electrically connected to the analog-to-digital converter input channel of the main control and computing unit for real-time phase current feedback. The thermistor is mounted on the surface of the coil of the three-phase magnetohydrodynamic thruster. The two ends of the thermistor are connected to another analog-to-digital converter input channel of the main control and computing unit for monitoring the coil temperature. The fault feedback pin of the three-phase full-bridge driver chip is electrically connected to the interrupt input pin of the main control and computing unit.

[0013] Preferably, the multimodal signal acquisition and processing unit includes a detection module, which comprises a single-photon avalanche diode detector array, a time-to-digital converter, a first level conversion chip, and a temperature stabilizer. The single-photon avalanche diode detector array includes sixteen independent detection pixel units. The avalanche signal output pin of each detection pixel unit is connected to the sixteen stop signal input channel pins of the time-to-digital converter. The start signal input channel pin of the time-to-digital converter is electrically connected to the synchronization signal output pin of the laser diode driver chip in the active disturbance generation unit to receive the excitation event start signal. The serial peripheral interface clock pin, serial peripheral interface master input slave output pin, and serial peripheral interface serial peripheral interface master input slave output pin of the time-to-digital converter are connected to the serial peripheral interface master input slave output pin and the serial peripheral interface master input slave output pin. The serial peripheral interface host output slave input pin is electrically connected to the serial peripheral interface host port of the main control and computing unit via a first level conversion chip for configuration and data reading. The interrupt request output pin of the time-to-digital converter is electrically connected to the external interrupt input pin of the main control and computing unit to indicate data readiness. The temperature stabilizer is mounted on the package of the single-photon avalanche diode detector array. The control signal input terminal of the temperature stabilizer is electrically connected to the general-purpose input / output port of the main control and computing unit to receive temperature control commands to maintain the stable operating temperature of the single-photon avalanche diode detector array. The reference clock input pin of the time-to-digital converter is electrically connected to the high-frequency reference clock signal provided by the communication and clock synchronization unit.

[0014] Preferably, the multimodal signal acquisition and processing unit further includes a flow field measurement module. The flow field measurement module includes an image sensor chip, a field-programmable gate array (FPGA) chip, a first temperature sensor, and a first memory. The pixel data output pin of the image sensor chip is connected to the mobile industrial processor interface (MIM) receiver pin of the FPGA chip via a mobile industrial processor interface channel. The synchronization signal input pin of the image sensor chip is electrically connected to the synchronization signal output pin of the three-phase full-bridge drive chip in the active disturbance generation unit to receive the flow field generation trigger signal. The control interface of the image sensor chip is electrically connected to the integrated circuit bus main controller pin of the FPGA chip to receive configuration commands. The chip is equipped with an image preprocessing accelerator intellectual property core and a particle image velocimetry calculation intellectual property core. The high-speed transceiver pin of the field-programmable gate array (FPGA) chip is electrically connected to the high-speed serial expansion interface of the main control and computing unit through a serial deserializer link to transmit the processed flow field vector data. The general-purpose input / output port of the FPGA chip is electrically connected to the control and data bus of the first memory for buffering image frames. The first temperature sensor is mounted on the package surface of the image sensor chip. The analog signal output pin of the first temperature sensor is connected to the analog-to-digital converter input channel of the main control and computing unit. The global clock input pin of the FPGA chip is electrically connected to the low-jitter differential clock signal provided by the communication and clock synchronization unit.

[0015] Preferably, the multimodal signal acquisition and processing unit further includes a spectral detection module, which includes a Raman laser driver, a spectrometer charge-coupled device (CCD), a dedicated spectral processing chip, a high-resolution analog-to-digital converter (ADC), a thermoelectric cooler, and a second temperature sensor. The analog dimming control pin of the Raman laser driver is electrically connected to the output pin of the ADC of the main control and computing unit to receive a light intensity control voltage. The laser enable pin of the Raman laser driver is electrically connected to the general-purpose input / output port of the main control and computing unit. The synchronous output pin of the Raman laser driver is electrically connected to the external sampling trigger pin of the high-resolution ADC. The photosensitive surface of the CCD receives Raman scattered light signals from the water body. The analog video signal output pin of the CCD is connected to the positive and negative analog signal input pins of the high-resolution ADC. The serial peripheral of the high-resolution ADC... The slave port of the interface is electrically connected to the serial peripheral interface host port of the dedicated spectral processing chip to transmit digitized spectral data. The dedicated spectral processing chip integrates hardware acceleration units for spectral accumulation and averaging, background subtraction, and baseline correction. Its high-speed serial data output pin is connected to the high-speed serial expansion interface of the main control and computing unit through the serial deserializer physical layer. The cold end of the thermoelectric cooler is mounted on the back of the charge-coupled device package of the spectrometer, and its drive signal input terminal is electrically connected to the pulse width modulation output pin of the dedicated spectral processing chip. The second temperature sensor is mounted on the surface of the charge-coupled device package of the spectrometer, and the output pin of the second temperature sensor is connected to the analog-to-digital converter input channel of the dedicated spectral processing chip. The over-temperature alarm pin of the dedicated spectral processing chip is electrically connected to the interrupt input pin of the main control and computing unit. The reference clock input pin of the Raman laser driver is electrically connected to the low phase noise reference clock signal provided by the communication and clock synchronization unit.

[0016] The present invention has the following beneficial effects: This invention applies controlled physical or chemical disturbances to water bodies, collects and analyzes their multi-dimensional response signals in real time, dynamically inverts the location and diffusion trend of pollution sources, and performs closed-loop learning after the task is completed to optimize subsequent performance. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a structural block diagram of the present invention.

[0019] 1. Main control and computing unit; 2. Active disturbance generation unit; 201. Quantum coherent excitation module; 202. Controllable chemical disturbance module; 203. Synthetic turbulent flow field module; 3. Multimodal signal acquisition and processing unit; 301. Detection module; 302. Flow field measurement module; 303. Spectroscopic detection module; 4. Power supply and energy management unit; 5. Communication and clock synchronization unit. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0021] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0022] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0023] In the description of this invention, it should be understood that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product of this invention is in use, or the orientation or positional relationship commonly understood by those skilled in the art. They are only used to facilitate the description of this invention and to simplify the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0024] Furthermore, the terms "first," "second," and "third" are used only to distinguish descriptions and should not be interpreted as indicating or implying relative importance.

[0025] In the description of this invention, it should also be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," and "link" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0026] A mobile water quality monitoring system, such as Figure 1 As shown, the system includes a main control and computing unit 1, an active disturbance generation unit 2, a multimodal signal acquisition and processing unit 3, a power supply and energy management unit 4, and a communication and clock synchronization unit 5. The active disturbance generation unit 2 includes a quantum coherent excitation module 201, a controllable chemical disturbance module 202, and a synthetic turbulence field module 203. The multimodal signal acquisition and processing unit 3 includes a detection module 301, a flow field measurement module 302, and a spectral detection module 303. The data bus interface of the main control and computing unit 1 is electrically connected to the data output interface of the multimodal signal acquisition and processing unit 3. The control signal port of the main control and computing unit 1 is electrically connected to the controlled end of the active disturbance generation unit 2. The data interface of the communication and clock synchronization unit 5 is electrically connected to the communication interface of the main control and computing unit 1. The first regulated output terminal, the second regulated output terminal, and the third regulated output terminal of the power supply and energy management unit 4 are electrically connected to the power input terminals of the main control and computing unit 1, the active disturbance generation unit 2, and the multimodal signal acquisition and processing unit 3, respectively. The main control and computing unit 1 is used to coordinate and execute intelligent decision-making and control throughout the entire process of environmental calibration, strategy generation, data inversion, decision scheduling, and model learning; the active disturbance generation unit 2 is used to receive and execute instructions from the main control unit and apply customized quantum optics, specific chemical tracers, and structured turbulent fields to the water body, as well as other multimodal physicochemical disturbances; the multimodal signal acquisition and processing unit 3 is used to synchronously acquire multidimensional response signals such as quantum state correlation, flow field structure distortion, and spectral fingerprint of the water body with disturbance events, and to perform real-time preprocessing and feature extraction; the power supply and energy management unit 4 is used to provide power to each unit and integrate energy recovery; the communication and clock synchronization unit 5 is responsible for time synchronization and data communication to ensure that instructions and data between distributed units are consistent in time and space.

[0027] The quantum coherent excitation module 201 includes a laser diode driver chip (Analog Devices LT3471), a temperature control chip (Analog Devices MAX1978ETJ+), a nonlinear optical crystal, a laser diode, a thermistor, and a thin-film heater (Minco). The HK9102R12.6L12B laser diode driver chip has its enable pin electrically connected to the general-purpose input / output port of the main control and computing unit 1 to receive the enable control signal. The feedback pin of the laser diode driver chip is grounded through the first sampling resistor. The switch output pin of the laser diode driver chip is connected to the anode of the laser diode, and the cathode of the laser diode is grounded. A nonlinear optical crystal is used to generate entangled photon pairs under the action of the pump light emitted by the laser diode. A thermistor is attached to the surface of the nonlinear optical crystal, and its two ends are respectively connected to the positive and negative input pins of the temperature sensor signal of the temperature control chip. A thin-film heater is attached to the surface of the nonlinear optical crystal, with one end of the thin-film heater connected to the power supply and the other end connected to the heater drive pin of the temperature control chip. The communication interface of the temperature control chip is electrically connected to the serial communication port of the main control and computing unit 1 to receive the temperature setpoint and return the temperature status. The synchronization signal output pin of the laser diode driver chip is electrically connected to the start channel input pin of the time-to-digital converter in the multi-mode signal acquisition and processing unit 3 to provide the excitation event start timestamp.

[0028] The controllable chemical perturbation module 202 includes a first motor driver chip (Texas Instruments DRV8873HPWPR), a high-voltage pulse generator chip (Maxim Integrated MAX038CPP), a micro-injection pump, and a microcavity electroporation release device. The logic input pins of the first motor driver chip are electrically connected to the pulse width modulation output port and direction control pin of the main control and computing unit 1 to receive motion control commands. The power input pins of the first motor driver chip are electrically connected to the motor drive power output terminal provided by the power supply and energy management unit 4. The bridge output pins of the first motor driver chip are connected to the two winding terminals of the DC motor in the micro-injection pump to drive its piston to reciprocate and inject or extract liquid tracer. The external trigger pin of the high-voltage pulse generator chip is electrically connected to the general-purpose input / output port of the main control and computing unit 1 to receive discharge trigger signals. The high-voltage power supply pin of the high-voltage pulse generator chip is connected to a high-voltage DC power supply through a current-limiting resistor. The ground pin of the high-voltage pulse generator chip is grounded. The high-voltage output pin of the high-voltage pulse generator chip is connected to the high-voltage electrode of the microcavity electroporation release device. The ground electrode of the microcavity electroporation release device is grounded. The internal cavity of the microcavity electroporation release device is used to contain capsules encapsulated with engineered microorganisms and generate a strong electric field when receiving a high-voltage pulse to break down the capsule membrane to achieve controlled release. The fluid outlet of the micro-injection pump is used to connect to the injection nozzle located in the water body through a pipeline. The fault status output pin of the first motor drive chip is electrically connected to the interrupt input pin of the main control and computing unit 1 for feedback of the drive status.

[0029] The synthetic turbulence field module 203 includes a three-phase full-bridge driver chip Infineon IR2136PBF, first to sixth power MOSFETs, first to third bootstrap diodes, first to third bootstrap capacitors, a three-phase magnetohydrodynamic thruster, and a current sensing chip from Texas Instruments. The INA240A2PWR and thermistor, along with the three high-side logic input pins and three low-side logic input pins of the three-phase full-bridge driver chip, are electrically connected to the pulse width modulation signal output port of the main control and computing unit 1 to receive six independent control waveforms. The three high-side gate drive pins of the three-phase full-bridge driver chip are connected to the gates of the first to third power MOSFETs through the first to third gate resistors, respectively. The three low-side gate drive pins of the three-phase full-bridge driver chip are connected to the gates of the fourth to sixth power MOSFETs through the fourth to sixth gate resistors, respectively. The anodes of the first to third bootstrap diodes are connected to the drive power supply, and their cathodes are connected to the first to third high-side floating power supply pins of the three-phase full-bridge driver chip, respectively. One end of the first to third bootstrap capacitors is connected to the first to third high-side floating power supply pins of the three-phase full-bridge driver chip, respectively, and the other end of the first to third bootstrap capacitors is connected to the first to third high-side reference pins of the three-phase full-bridge driver chip, respectively. The voltage pins, the first to third high-side reference voltage pins are electrically connected to the sources of the fourth to sixth power MOSFETs respectively. The drains of the first to third power MOSFETs are connected to the high-voltage DC bus. The sources of the first to third power MOSFETs serve as the three-phase output terminals respectively. The drains of the fourth to sixth power MOSFETs are electrically connected to the three-phase output terminals respectively. The sources of the fourth to sixth power MOSFETs are connected to the current detection pin of the current detection chip and grounded. The three-phase output terminals are connected to the three coil input terminals of the three-phase magnetohydrodynamic thruster respectively. The voltage output pin of the current detection chip is electrically connected to the analog-to-digital converter input channel of the main control and computing unit 1 for real-time phase current feedback. The thermistor is mounted on the surface of the coil of the three-phase magnetohydrodynamic thruster. The two ends of the thermistor are connected to another analog-to-digital converter input channel of the main control and computing unit 1 for monitoring the coil temperature. The fault feedback pin of the three-phase full-bridge drive chip is electrically connected to the interrupt input pin of the main control and computing unit 1.

[0030] Detection module 301 includes an Excelitas C30902SH-500 single-photon avalanche diode detector array, an ACAM Messelectronic TDC-GPX2 time-to-digital converter, a Texas Instruments TXS0108EPWR first-level conversion chip, and a TE Connectivity 1-1623939-3 temperature stabilizer. The single-photon avalanche diode detector array includes sixteen independent detection pixel units. The avalanche signal output pin of each detection pixel unit is connected to the sixteen stop signal input channel pins of the time-to-digital converter. The start signal input channel pin of the time-to-digital converter is electrically connected to the synchronization signal output pin of the laser diode driver chip in the active disturbance generation unit 2 to receive the excitation event start signal. The serial peripheral interface clock pin, serial peripheral interface master input slave output pin, and serial peripheral interface master output slave input pin of the time-to-digital converter are connected to the master controller and the main controller via the first-level conversion chip. The serial peripheral interface host port of computing unit 1 is electrically connected for configuration and data reading. The interrupt request output pin of the time-to-digital converter is electrically connected to the external interrupt input pin of the main control and computing unit 1 to indicate that the data is ready. The temperature stabilizer is mounted on the package of the single-photon avalanche diode detector array. The control signal input terminal of the temperature stabilizer is electrically connected to the general-purpose input / output port of the main control and computing unit 1 to receive temperature control commands to maintain the stable operating temperature of the single-photon avalanche diode detector array. The reference clock input pin of the time-to-digital converter is electrically connected to the high-frequency reference clock signal provided by the communication and clock synchronization unit 5.

[0031] The flow field measurement module 302 includes an Onsemi PYTHON 1300 image sensor chip, an Intel Cyclone 10 GX 10CX150 field-programmable gate array (FPGA) chip, an Analog Devices TMP117MAIDRVR temperature sensor, and a Micron MT40A256M16GE-083E memory. The pixel data output pins of the image sensor chip are connected to the FPGA's FPGA receiver pins via a mobile industrial processor interface channel. The image sensor chip's synchronization signal input pins are electrically connected to the synchronization signal output pins of the three-phase full-bridge driver chip in the active disturbance generation unit 2 to receive the flow field generation trigger signal. The image sensor chip's control interface is electrically connected to the FPGA's integrated circuit bus main controller pins to receive configuration commands. The FPGA chip is equipped with an image preprocessing accelerator core and a particle image velocimeter. The high-speed transceiver pins of the field-programmable gate array (FPGA) chip are electrically connected to the high-speed serial expansion interface of the main control and computing unit 1 via a serial deserializer link to transmit processed flow field vector data. The general-purpose input / output ports of the FPGA chip are electrically connected to the control and data bus of the first memory for buffering image frames. The first temperature sensor is mounted on the package surface of the image sensor chip. The analog signal output pin of the first temperature sensor is connected to the analog-to-digital converter input channel of the main control and computing unit 1. The global clock input pin of the FPGA chip is electrically connected to the low-jitter differential clock signal provided by the communication and clock synchronization unit 5.

[0032] The spectral detection module 303 includes a Raman laser driver (Analog Devices ADN8810BRUZ), a spectrometer charge-coupled device (CCD), a dedicated spectral processing chip (Analog Devices ADDI9036), a high-resolution analog-to-digital converter (Texas Instruments ADS127L11IPBSR), a thermoelectric cooler, and a second temperature sensor (Analog). The Devices ADT7310TRZ Raman laser driver's analog dimming control pin is electrically connected to the output pin of the digital-to-analog converter (DAC) of the main control and computing unit 1 to receive the light intensity control voltage. The Raman laser driver's laser enable pin is electrically connected to the general-purpose input / output port of the main control and computing unit 1. The Raman laser driver's synchronous output pin is electrically connected to the external sampling trigger pin of the high-resolution DAC. The photosensitive surface of the spectrometer's charge-coupled device (CCD) receives the Raman scattered light signal from the water body. The analog video signal output pin of the CCD is connected to the analog signal input positive and negative pins of the high-resolution DAC. The serial peripheral interface slave port of the high-resolution DAC is electrically connected to the serial peripheral interface master port of the dedicated spectral processing chip to transmit the digitized light. The dedicated spectral processing chip integrates hardware acceleration units for spectral accumulation and averaging, background subtraction, and baseline correction. Its high-speed serial data output pin is connected to the high-speed serial expansion interface of the main control and computing unit 1 through the serial deserializer physical layer. The cold end of the thermoelectric cooler is mounted on the back of the charge-coupled device package of the spectrometer, and its drive signal input terminal is electrically connected to the pulse width modulation output pin of the dedicated spectral processing chip. The second temperature sensor is mounted on the surface of the charge-coupled device package of the spectrometer, and the output pin of the second temperature sensor is connected to the analog-to-digital converter input channel of the dedicated spectral processing chip. The over-temperature alarm pin of the dedicated spectral processing chip is electrically connected to the interrupt input pin of the main control and computing unit 1. The reference clock input pin of the Raman laser driver is electrically connected to the low phase noise reference clock signal provided by the communication and clock synchronization unit 5.

[0033] The main control and computing unit 1 specifically includes a multi-core microprocessor NXP i.MX 8M Plus, a dynamic Bayesian inversion coprocessor Xilinx Versal AI Core VC1902, a system memory chip Micron MT40A512M16LY-075E, a non-volatile memory chip Kioxia TH58LJT0T24BSA8, a high-speed serial computer expansion bus standard interface switching chip Microchip PM8536, a first Ethernet physical layer chip Microchip LAN8742A, and a real-time clock chip MaximIntegrated. The DS3231MZ+ multi-core microprocessor's multiple general-purpose input / output ports are electrically connected to the enable pins of the laser diode driver chip, the logic input pins of the first motor driver chip, the logic input pins of the three-phase full-bridge driver chip, and the enable pin of the load switch chip in the power supply and energy management unit 4, respectively. The first set of high-speed serial computer expansion bus standard channels of the multi-core microprocessor is electrically connected to the upstream port of the high-speed serial computer expansion bus standard interface switching chip. Multiple downstream ports of the switching chip are electrically connected to the high-speed serial computer expansion bus standard interfaces of the field-programmable gate array chip and the dedicated spectral processing chip in the multi-mode signal acquisition and processing unit 3, respectively. The multi-core microprocessor's gigabit media independent interface is electrically connected to the first Ethernet physical layer chip and then to the Ethernet switch in the communication and clock synchronization unit 5 via an Ethernet line. The multi-core microprocessor's dynamic random access memory interface is electrically connected to the system memory chip. The multi-core microprocessor's serial peripheral interface... The host port is electrically connected to the serial peripheral interface slave port of the time-to-digital converter chip and the high-resolution analog-to-digital converter chip, respectively. The multiple analog-to-digital converter input channels of the multi-core microprocessor are electrically connected to the voltage output pin of the current detection chip and the output pins of the first and second temperature sensors, respectively. The dynamic Bayesian inversion coprocessor is directly connected to the memory controller of the multi-core microprocessor through a high-speed memory bus to accelerate particle filtering and causal discovery algorithms. The non-volatile memory chip is connected to the multi-core microprocessor through a serial peripheral interface or an embedded multimedia card interface to store system programs, model parameters, and task data. The serial communication interface of the real-time clock chip is connected to the integrated circuit bus of the multi-core microprocessor, and its interrupt output pin is electrically connected to the external interrupt input pin of the multi-core microprocessor to provide a system timing reference. The universal asynchronous transceiver port of the multi-core microprocessor is electrically connected to the serial data port of the 5G communication module, the serial data port of the GPS disciplined clock module, and the communication interface of the temperature control chip, respectively.

[0034] Power and energy management unit 4 includes a main buck converter chip (Texas Instruments LM5176RHFT), first to third load switch chips (Texas Instruments TPS25940AAPWR), energy harvesting management chip (AnalogDevices LTC3108EDE), and voltage and current monitoring chip (Texas Instruments). The system includes an INA228AIDGSR, a supercapacitor bank, and a backup lithium battery bank. The voltage input pin of the main buck converter chip is electrically connected to the positive terminal of the external 48V DC power supply. The feedback pin of the main buck converter chip is electrically connected to the output voltage terminal through a series voltage divider network of the first and second precision voltage dividers. The switching output pin of the main buck converter chip is electrically connected to the output voltage terminal through a power inductor to step down the input voltage to the system's main 12V power rail. The system's main 12V power rail is connected to the power input pins of the first, second, and third load switch chips, respectively. The enable pin of the first load switch chip is electrically connected to the first general-purpose input / output port of the main control and computing unit 1, and its power output pin is connected to the power input pin of the active disturbance generation unit 2. The enable pin of the second load switch chip is electrically connected to the second general-purpose input / output port of the main control and computing unit 1, and its power output pin is connected to the power input pin of the multi-mode signal acquisition and processing unit 3. The source input pin and the enable pin of the third load switch chip are electrically connected to the third general-purpose input / output port of the main control and computing unit 1. Its power output pin is connected to the power input pin of the communication and clock synchronization unit 5. The positive and negative terminals of the current detection input pin of the voltage and current monitoring chip are connected in series to the main 12V power rail of the system. Its voltage detection input pin is electrically connected to the main 12V power rail of the system. Its data interface is connected to the main control and computing unit 1 through the integrated circuit bus and is used to report voltage, current and power parameters. The AC input pin 1 and input pin 2 of the energy harvesting management chip are connected to the two-phase output terminals of the micro hydroelectric generator. Its DC output pin is connected to the positive terminal of the supercapacitor bank. The negative terminal of the supercapacitor bank is grounded to form an energy recovery and storage branch. The positive terminal of the backup lithium battery bank is electrically connected to the main 12V power rail of the system through a diode. Its negative terminal is grounded to provide seamless backup power when the main power is interrupted. The status indicator pin of the energy harvesting management chip is electrically connected to the interrupt input pin of the main control and computing unit 1.

[0035] The communication and clock synchronization unit 5 includes a Global Positioning System disciplined clock module (Microchip SA.65s), a fifth-generation mobile communication technology module (Quectel RM520N-GL), a multi-port gigabit Ethernet switch chip (Microchip KSZ9897RTXI), a fiber optic transceiver module (Intel E10G41BTDA), and a voltage regulator chip (Texas). The Instruments TPS7A4701RGWR and antenna assembly, along with the pulse-per-second output pin of the GPS-disciplined clock module, are connected to the external interrupt input pin of the multi-core microprocessor in the main control and computing unit 1 to provide a unified millisecond-level time synchronization reference for the entire system. Its serial time data output pin is electrically connected to the universal asynchronous transceiver (UART) receive pin of the multi-core microprocessor to transmit messages containing Coordinated Universal Time (UTC) information. The UART transmit and receive pins of the fifth-generation mobile communication (5G) module are electrically connected to another set of UART receive and transmit pins of the multi-core microprocessor, respectively. Its main antenna port is connected to the external 5G external antenna in the antenna assembly via a coaxial cable. The uplink media independent interface of the multi-port gigabit Ethernet switch chip is connected to the main control and computing unit 1 through the first Ethernet physical layer chip, and its multiple downlink media independent interfaces... The interfaces are connected to the local controllers of the field-programmable gate array chip and the spectrum processing chip in the multi-mode signal acquisition and processing unit 3 through the corresponding Ethernet physical layer chips to form a distributed high-speed data network. The electrical signal interface of the fiber optic transceiver module is connected to another downlink media independent interface of the multi-port gigabit Ethernet switch chip. Its optical signal interface is connected to the remote command center through optical fiber to provide a long-distance, interference-resistant data backhaul link. The input end of the voltage regulator chip is electrically connected to the communication unit power rail provided by the power supply and energy management unit 4. Its output end provides the required stable low-voltage operating power to the GPS disciplined clock module, the fifth-generation mobile communication technology module, the multi-port gigabit Ethernet switch chip, and the fiber optic transceiver module. The antenna assembly also includes a GPS antenna, whose signal output end is connected to the antenna input pin of the GPS disciplined clock module through a coaxial cable.

[0036] The above system operates through the following steps: S1: System initialization and collaborative calibration steps: The main control and computing unit 1 controls the power supply and energy management unit 4 to power on each unit, and calibrates the clock of the entire system through the communication and clock synchronization unit 5; the main control and computing unit 1 instructs the active disturbance generation unit 2 to remain silent, and controls the multimodal signal acquisition and processing unit 3 to acquire background environmental parameters and establish baseline models for each sensor. S1.1: The main control and computing unit 1 receives the pulse signal per second and serial time code sent by the high-precision clock source through the communication and clock synchronization unit 5, thereby synchronizing its internal clock and the clock reference distributed to the active disturbance generation unit 2 and the multi-mode signal acquisition and processing unit 3 through the communication network. S1.2: When the active disturbance generation unit 2 is not working, the main control and computing unit 1 controls the quantum detection module 301, flow field measurement module 302 and spectral detection module 303 in the multimodal signal acquisition and processing unit 3 to continuously acquire background signals of the monitored water area for a period of time. S1.3: The main control and computing unit 1 performs statistical analysis on the collected background signals, calculates the average value, variance and cross-correlation of each signal channel, and establishes a multivariate statistical baseline model for subsequent anomaly detection and signal separation. S2: Adaptive perturbation strategy generation and scheduling steps: Based on the current environmental baseline, historical data and suspected pollution information, the main control and computing unit 1 calculates the multimodal perturbation scheme with the maximum expected information gain; the main control and computing unit 1 generates a precise instruction set containing temporal and spatial coordinates, and sends it to the quantum excitation, chemical injection and turbulence generation modules in the active perturbation generation unit 2 through the control bus respectively; S2.1: The main control and computing unit 1 constructs one or more initial hypotheses about potential pollution sources based on the hydrological data and historical pollution event database transmitted back in real time by the multimodal signal acquisition and processing unit 3; S2.2: For each initial hypothesis, the main control and computing unit 1 simulates and evaluates various disturbance forms and parameter combinations generated by different actuators in the active disturbance generation unit 2, including lasers, injection pumps, and magnetohydrodynamic thrusters, and predicts the response signals that may be triggered. S2.3: The main control and computing unit 1 uses the Bayesian experimental design criterion to calculate the expected information gain that each simulation perturbation scheme can provide for distinguishing different initial assumptions, and selects the scheme with the largest expected information gain globally. S2.4: The main control and computing unit 1 decomposes the selected optimal disturbance scheme into a specific sequence of equipment control instructions, including the enable timing of the laser driver chip, the start / stop and flow rate of the injection pump, and the modulation waveform parameters of the three-phase full-bridge driver chip. S3: Multimodal disturbance execution and synchronous response acquisition steps: Active disturbance generation unit 2 receives and executes the instruction set to apply controlled physical or chemical disturbances to the target water body; at the same time, the main control and computing unit 1 triggers the multimodal signal acquisition and processing unit 3 to acquire quantum optical, hydrodynamic and chemical / biological response signals of the water body in strict synchronization with the disturbance event; S3.1: The laser diode driver chip in the active perturbation generation unit 2 drives the laser diode to emit entangled photon pulses of a specific pulse width after its enable pin is set according to the instruction. S3.2: The motor drive chip in the active disturbance generation unit 2 drives the micro injection pump to inject tracer at a set rate under the control of the pulse width modulation signal received at its logic input pin according to the instruction; S3.3: The three-phase full-bridge drive chip in the active disturbance generation unit 2 receives the modulation waveform on its three pulse width modulation signal input pins according to the instruction, and drives the magnetohydrodynamic thruster to generate a synthetic turbulent field with a specific structure. S3.4: At the same moment when each disturbance action begins, the main control and computing unit 1 sends a synchronous acquisition trigger signal to the multimodal signal acquisition and processing unit 3; S3.5: The time-to-digital converter chip in the multimodal signal acquisition and processing unit 3 starts timing after receiving a trigger on its start signal input channel pin, and records the timestamp of the photon event transmitted by the single-photon avalanche diode detector array; at the same time, the high frame rate image sensor and the Raman spectrometer start acquiring flow field images and spectral signals synchronously. S4: Data fusion and dynamic source tracing inversion steps: The multimodal signal acquisition and processing unit 3 preprocesses the acquired raw data and transmits it to the main control and computing unit 1; the main control and computing unit 1 runs the data fusion algorithm and dynamic Bayesian inversion model, combines multi-source heterogeneous data with the prior hydrodynamic model, and iteratively updates the posterior probability distribution of pollution source location, emission intensity and pollution cloud diffusion range. S4.1: The field-programmable gate array chip in the multimodal signal acquisition and processing unit 3 performs real-time cross-correlation calculation on the image sensor data and extracts the flow field vector data; the time-to-digital converter chip uploads the coincidence count histogram data through the serial peripheral interface; and the high-resolution analog-to-digital converter chip uploads the spectral data. S4.2: The main control and computing unit 1 receives the above data and executes the quantum tomography algorithm to invert the perturbation optical transmission matrix of the water body from the coincidence count histogram; executes the particle image velocimetry post-processing algorithm to parse the distortion field of the synthetic turbulence from the flow field vector data; and executes the spectral deconvolution algorithm to identify the fingerprint of characteristic pollutants from the spectral data. S4.3: The main control and computing unit 1 combines the extracted multidimensional feature vectors with the prior hydrodynamic model to construct a state space model; it adopts a particle filtering algorithm based on sequential importance sampling to approximate the joint posterior probability density function of pollution source parameters and pollution field by continuously predicting and updating particle states and their weights. S5: System learning and model update steps: The main control and computing unit 1 uses the full-chain data of this task to adaptively optimize and update the built-in hydrodynamic model parameters, sensor noise model and disturbance strategy model, and stores the updated model to improve the performance of subsequent monitoring tasks. S5.1: After the task is completed, the main control and computing unit 1 summarizes the timing data of the entire process from initialization to handling verification, including disturbance commands, multimodal responses, inversion intermediate results and final handling effect feedback; S5.2: The main control and computing unit 1 optimizes the model prediction accuracy by solving a maximum a posteriori probability estimation problem and using the collected data to perform reverse calibration on key parameters in the hydrodynamic-water quality coupling model, including the diffusion coefficient and attenuation coefficient. S5.3: The main control and computing unit 1 analyzes the noise performance of each sensor in this mission and updates the noise covariance matrix estimate in the multivariate statistical baseline model; S5.4: The main control and computing unit 1 uses the successful perturbation strategies and inversion results from this task as new samples and stores them in the strategy knowledge base to enrich the prior knowledge of adaptive perturbation strategy generation in future step S2.

[0037] This mobile water quality monitoring system addresses the problems of model cognitive bias and emergency resource mismatch caused by data spatiotemporal sparsity in traditional mobile monitoring systems when dealing with intermittent illegal discharges and dynamic hydrological conditions.

[0038] Upon receiving an alarm about an anomaly in the downstream river of the industrial park, the main control and computing unit 1 initiated and coordinated a rapid response across the entire system.

[0039] First, the main control unit reads the precise time message sent by the GPS-disciplined clock module in the communication and clock synchronization unit 5 through its communication interface, and captures its pulse signal per second, which is used as the absolute time reference for the entire system. The synchronization clock signal is then distributed to all distributed units through a network composed of Ethernet switch chips and fiber optic transceiver modules to ensure the synchronization accuracy of all subsequent actions.

[0040] This initial synchronization action laid a solid foundation for solving the problem of poor spatiotemporal correlation of events under dynamic hydrological conditions.

[0041] At the same time, the power supply and energy management unit 4 starts to work. Its main buck converter chip efficiently converts the external main power supply into the system main power rail, and the load switch chip, under the logic control of the main control unit, sequentially powers on the active disturbance generation unit 2, the multi-mode signal acquisition and processing unit 3, etc.

[0042] Among them, the energy harvesting management chip continuously obtains energy from the micro hydroelectric generator and charges the supercapacitor bank, while the voltage and current monitoring chip reports real-time power consumption to the main control unit to ensure stable operation in the field for a long time.

[0043] Next, the system performs initialization and collaborative calibration steps. At this time, the active disturbance generation unit 2 remains silent, while the multimodal signal acquisition and processing unit 3 begins to acquire the environmental background.

[0044] In the quantum detection module 301, the single-photon avalanche diode detector array begins to capture ambient background photons under the control of a temperature stabilizer, and its output avalanche signal is recorded by a time-to-digital converter chip.

[0045] In the flow field measurement module 302, the image sensor chip is driven by the field programmable gate array chip to capture background flow field images under the monitoring of the temperature sensor, and caches them in the memory chip.

[0046] In the spectral detection module 303, the spectrometer charge-coupled device, cooled by a thermoelectric cooler, acquires the background Raman spectrum of the water body through a high-resolution analog-to-digital converter chip, and performs preliminary processing by a dedicated spectral processing chip.

[0047] All this background data is aggregated to the main control unit via a high-speed serial bus switching chip or Ethernet network. The multi-core microprocessor, in conjunction with a coprocessor, performs statistical analysis to establish a multi-dimensional environmental baseline model, including the mean photon count rate, flow velocity distribution, and spectral characteristic peaks, and stores it in a non-volatile memory chip.

[0048] The effect of this step is to provide a pure zero-value reference for subsequent identification of abnormal signals, effectively removing environmental background interference, making the signal distortion caused by weak pollution more prominent, and directly addressing the shortcomings of traditional methods that cannot identify low-concentration or transient pollution signals due to background noise masking.

[0049] Faced with the challenges of fleeting pollution traces and dynamically changing hydrological conditions in the downstream river channels of industrial parks, the adaptive disturbance strategy generation and scheduling steps begin to play a role.

[0050] Based on the established environmental baseline, potentially accessible real-time hydrological data, and suspicious areas in alarm information, the main control and computing unit 1 utilizes the powerful parallel computing capabilities of the coprocessor to run an optimization algorithm based on Bayesian experimental design.

[0051] The algorithm simulates the differences in expected information gain caused by different perturbation methods and their various parameter combinations, such as quantum coherent excitation, injection of specific tracers, and synthetic turbulent fields.

[0052] The core of expected information gain is to assess the extent to which different perturbation schemes can reduce the posterior uncertainty of unknown parameters such as the location, emission intensity, and time of potential pollution sources.

[0053] The algorithm rapidly simulates the potential spread range of pollution plumes in a dynamic hydrological model and predicts the theoretical response of different disturbances in the simulated environment, thereby selecting the most effective optimization strategy from a vast number of possibilities to illuminate potential pollution areas and obtain the most discriminative information.

[0054] Subsequently, the main control unit converts the strategy into a control command sequence accurate to the millisecond level. It sends a high-level command to the enable pin of the laser diode driver chip in the active disturbance generation unit 2 through its general-purpose input / output port. At the same time, it sends the set value to the temperature control chip through serial communication. The latter drives the thin-film heater and reads the resistance value of the thermistor to precisely maintain the nonlinear optical crystal at the phase-matching temperature, ensuring that the pump light energy emitted by the laser diode efficiently generates entangled photons.

[0055] The purpose of this quantum coherent excitation perturbation is to obtain material fingerprint information that cannot be provided by traditional optical methods through the interaction of photons with the quantum state of pollutant molecules.

[0056] Through its pulse width modulation port and direction control pin, it sends control waveforms to the motor drive chip in the controllable chemical perturbation module 202 to drive the DC motor in the micro-injection pump to inject the selected chemical tracer or engineered microbial capsule carrier into the water.

[0057] Through another set of pulse width modulation ports, six independent modulation waveforms are sent to the three-phase full-bridge drive chip in the synthetic turbulence field module 203. The driver controls the on and off of the power transistor through the gate resistor. Combined with the floating power supply composed of bootstrap diodes and capacitors, the three-phase magnetohydrodynamic thruster coil generates a vortex array of specific size and intensity under the drive of the high-voltage DC bus.

[0058] This synthetic turbulence was used to actively mix water bodies and revealed changes in local rheological properties caused by pollutants.

[0059] Instead of passively waiting and randomly capturing pollutant plumes that may have been diluted or migrated, it actively injects a series of known and controllable physicochemical probes into complex and dynamic aquatic systems.

[0060] These probes interact with pollutants that may be present in the water through specific quantum state scattering, chemical complexation, or hydrodynamic interactions, thus transforming the problem of "finding random sparse pollution patches" into a high signal-to-noise ratio observation problem of "detecting whether a known disturbance signal undergoes predictable pattern distortion".

[0061] It overcomes the data sparsity caused by the intermittency of illegal discharge and the randomness of monitoring, and provides multi-dimensional correlation features with clear physical meaning for subsequent data fusion.

[0062] This is followed by the multimodal disturbance execution and synchronous response acquisition steps.

[0063] At the same moment when all disturbances are initiated according to the predetermined timing, the main control unit sends a hardware synchronization trigger signal to the multi-mode signal acquisition and processing unit 3.

[0064] In the quantum detection module 301, the start channel input pin of the time-to-digital converter chip receives a trigger signal from the synchronous output pin of the laser driver, and begins to record the timestamps of the photon stop signals that return after the representative signal light from each pixel of the single-photon avalanche diode array has been transmitted in the water.

[0065] Subsequent coincidence counting and quantum tomography algorithms can be used to inversely determine the subtle changes in the optical transport matrix of water caused by pollutants, changes that are extremely sensitive to specific molecular structures.

[0066] In the flow field measurement module 302, the synchronization input pin of the image sensor chip receives a synchronization signal from the driver chip and begins to capture massive amounts of pixel data generated by the disturbance of the synthetic turbulent flow field and the motion of tracer particles at a high frame rate.

[0067] The data is transmitted in real time to a field-programmable gate array (FPGA) chip via a high-speed interface. The chip then performs real-time cross-correlation calculations using its internally embedded image preprocessing and particle image velocimetry calculation core, outputting a high spatiotemporal resolution two-dimensional or three-dimensional velocity vector field.

[0068] Any local density or viscosity change caused by pollutants will lead to abnormal attenuation, deflection, or vortex structure distortion in the synthetic turbulence.

[0069] In the spectral detection module 303, the Raman laser driver emits excitation light of a specific power after receiving an enable signal, and its synchronous output pin triggers the analog-to-digital converter to start sampling the Raman scattered light generated by the water body captured by the charge-coupled device.

[0070] The spectral fingerprint of this scattered light will exhibit characteristic peak shifts or intensity changes due to the presence of contaminants.

[0071] The digitized spectral data is then processed in real time by a dedicated spectral processing chip, which performs cumulative averaging, background subtraction, and baseline correction to extract weak feature signals.

[0072] By utilizing multimodal active perturbation as a coordinated structured illumination and excitation probe, a comprehensive, multi-physical, and strictly synchronized tomographic scan or excitation response spectrum measurement was performed on dynamically changing water bodies.

[0073] What was obtained was not a simple concentration scalar data of a few discrete points, but a continuous spatiotemporal dataset containing multi-field coupled response characteristics such as the quantum correlation characteristics of the light field, the flow field structure function, and the vibrational spectra of matter molecules under the influence of pollutants.

[0074] Its data density, information dimensionality, and physical interpretability far exceed the sporadic snapshots obtained from traditional single-trip surveys, providing extremely rich and mutually verifying observational constraints for subsequent accurate inversion and solving the data sparsity problem.

[0075] After acquiring massive amounts of multimodal data, the system enters the data fusion and dynamic source tracing and inversion process.

[0076] After the preprocessing unit of each acquisition module adapts the level through the level conversion chip, it transmits the normalized feature data to the main control and computing unit 1 in real time via the high-speed data transmission channel built by the high-speed serial bus switching chip or gigabit Ethernet.

[0077] The multi-core microprocessor and coprocessor of the main control unit begin to work together to run a complex joint inversion algorithm.

[0078] The algorithm first performs feature decoding in parallel. From the quantum coincidence count data, it inverses the perturbation field of the water body's optical transport matrix through maximum likelihood estimation. This perturbation field implies information about pollutant concentration and composition. From the flow field vector data, it extracts the equivalent viscosity or density perturbation field caused by pollutants by solving the inverse hydrodynamic problem. From the Raman spectral data, it identifies the composition and relative concentration of specific pollutants through spectral library matching and deconvolution.

[0079] These decoded multidimensional feature fields serve as observational evidence and are closely integrated with a high-precision hydrodynamic-water quality coupling model that dynamically assimilates pre-loaded real-time hydrological monitoring data accessed via a communication unit.

[0080] The model runs on the coprocessor's computing engine and can simulate hydrodynamic and material transport processes in real time.

[0081] The system employs advanced Bayesian inference algorithms such as particle filtering based on sequential importance sampling, and is accelerated by coprocessor hardware.

[0082] The algorithm uses a large number of particles to assume different pollution source scenarios. Each particle contains a set of parameters such as the location of the pollution source, emission history, and intensity.

[0083] Each particle predicts the evolution of the pollution cloud in parallel according to the dynamic hydrodynamic model, and compares the predicted modal observations, namely the predicted optical disturbance field, flow field distortion field, and spectral features, with the actual decoded feature field in multiple dimensions and multiple physical quantities to calculate a comprehensive likelihood probability.

[0084] Particles that match well with all multimodal observation data, i.e., those with more reasonable pollution source hypotheses, have increased weights, and vice versa.

[0085] By continuously iterating, predicting, updating, and resampling the weighted set of all particles, the algorithm eventually converges to a dynamic posterior probability estimate of pollution source parameters, accurate to latitude and longitude coordinates and emission time curves, and the real-time three-dimensional distribution of pollution clouds and their future short-term diffusion paths. This changes the traditional simple extrapolation paradigm based on single sampling and fixed parameter models.

[0086] By deeply fusing multi-source heterogeneous, high-dimensional observational data with dynamic physical models within a rigorous probabilistic framework, the system can robustly deduce the most likely complete storyline of a pollution event from highly uncertain or even noisy observations. This includes the precise location of the pollution source, the start and end times and intensity changes of emissions, as well as the true three-dimensional morphology, concentration gradient, and movement trajectory of the pollution cloud in the current and future periods.

[0087] This generates a hologram of the pollution situation based on probability density rather than a single deterministic line, and which can be updated over time.

[0088] It corrects serious cognitive biases caused by data sparsity, distorted model parameters, such as the use of incorrect flow rates, enabling emergency decision-makers to gain a comprehensive understanding of pollution events, rather than just a partial view.

[0089] To further enhance the reliability and decision confidence of the inversion results, the system can then be integrated into an intelligent decision verification process. The main control and computing unit 1 can initiate an adversarial digital twin verification mechanism based on the posterior probability distribution obtained from the current inversion. This involves running two or more parallel digital twin environment models on the coprocessor simultaneously. One model accepts the mainstream hypothesis of the current posterior probability, such as the location of the pollution source with the highest probability, while another one or more adversarial models actively search for alternative hypotheses that have a similar overall fit to the existing observation data but differ significantly in the location or emission pattern of the pollution source.

[0090] Then, the system will use game theory to design a decisive experiment, which involves calculating a new small-scale active perturbation scheme that is expected to maximize the divergence in the predictions of the observation results among the competing hypotheses.

[0091] The main control unit then instructs the active disturbance generation unit 2 to execute this brief verification disturbance, and the multi-mode signal acquisition and processing unit 3 synchronously acquires the response.

[0092] By analyzing the new data from this targeted experiment, the system can confirm or rule out certain competing hypotheses, much like a decisive test in a scientific experiment. This can raise the confidence level of the final judgment on the pollution source to a near-confirmatory level, transforming emergency response from a gambling-style decision based on a single possibility and greatly reducing the risk of misjudgment due to model uncertainty or data randomness.

[0093] Finally, the entire event did not end with the conclusion of the source; the system will enter the system learning and model update steps to achieve continuous evolution.

[0094] The main control and computing unit 1 archives all data from the entire task chain, including initial environmental data, all active perturbation command sequences, raw and processed multimodal response data, inversion intermediate results and final probability distribution, as well as any possible verification results.

[0095] Using these valuable field experimental data, the system automatically calibrates and optimizes key localized parameters in the hydrodynamic-water quality coupling model, such as the lateral diffusion coefficient and pollutant degradation rate of the river section, by solving the maximum a posteriori probability estimation problem. At the same time, it updates the noise models of each sensor and the background baseline model, and stores the successful perturbation strategy and the inversion evidence chain as new cases in the strategy knowledge base.

[0096] This enables the system to learn from experience. Each task deepens its understanding of the coupling patterns of hydrological pollution in a specific water area, makes its model more accurate, and its strategies more effective. As a result, when faced with new intermittent pollution events in the future, it can complete source tracing and response more quickly and accurately, forming a virtuous cycle of becoming smarter with use.

[0097] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A mobile water quality monitoring system, characterized in that, include The main control and computing unit (1) is used to coordinate the intelligent decision-making and control of the entire process of environmental calibration, strategy generation, data inversion, decision scheduling and model learning. The active disturbance generation unit (2) is used to receive and execute instructions from the main control unit and apply customized quantum optics, specific chemical tracers and structured turbulent fields and other multimodal physical and chemical disturbances to the water body. The multimodal signal acquisition and processing unit (3) is used to synchronously acquire multidimensional response signals such as quantum state correlation, flow field structure distortion and spectral fingerprint of water body with disturbance events and perform real-time preprocessing and feature extraction. The power supply and energy management unit (4) is used to provide power to each unit and integrate energy recovery; The communication and clock synchronization unit (5) is responsible for time synchronization and data communication to ensure that instructions and data between distributed units are consistent in time and space.

2. The mobile water quality monitoring system according to claim 1, characterized in that, The data bus interface of the main control and computing unit (1) is electrically connected to the data output interface of the multimodal signal acquisition and processing unit (3). The control signal port of the main control and computing unit (1) is electrically connected to the controlled end of the active disturbance generation unit (2). The data interface of the communication and clock synchronization unit (5) is electrically connected to the communication interface of the main control and computing unit (1). The first regulated output terminal, the second regulated output terminal, and the third regulated output terminal of the power supply and energy management unit (4) are electrically connected to the power input terminals of the main control and computing unit (1), the active disturbance generation unit (2), and the multimodal signal acquisition and processing unit (3), respectively.

3. The mobile water quality monitoring system according to claim 1, characterized in that, The active perturbation generating unit (2) includes a quantum coherent excitation module (201), which includes a laser diode driver chip, a temperature control chip, a nonlinear optical crystal, a laser diode, a thermistor, and a thin-film heater. The enable pin of the laser diode driver chip is electrically connected to the general-purpose input / output port of the main control and computing unit (1) to receive an enable control signal. The feedback pin of the laser diode driver chip is grounded through a first sampling resistor. The switch output pin of the laser diode driver chip is connected to the anode of the laser diode. The cathode of the laser diode is grounded. The nonlinear optical crystal is used to generate entangled photon pairs under the action of the pump light emitted by the laser diode. The thermistor is attached to the surface of the nonlinear optical crystal. The two ends of the thermistor are respectively connected to the positive and negative input pins of the temperature sensor signal of the temperature control chip. The thin film heater is attached to the surface of the nonlinear optical crystal. One end of the thin film heater is connected to the power supply. The other end of the thin film heater is connected to the heater drive pin of the temperature control chip. The communication interface of the temperature control chip is electrically connected to the serial communication port of the main control and computing unit (1) to receive the temperature set value and return the temperature status. The synchronization signal output pin of the laser diode driver chip is electrically connected to the start channel input pin of the time-to-digital converter in the multimodal signal acquisition and processing unit (3) to provide the start timestamp of the excitation event.

4. The mobile water quality monitoring system according to claim 1, characterized in that, The active perturbation generating unit (2) further includes a controllable chemical perturbation module (202), which includes a first motor drive chip, a high-voltage pulse generating chip, a micro-injection pump, and a microcavity electroporation release device. The logic input pin of the first motor drive chip is electrically connected to the pulse width modulation output port and direction control pin of the main control and computing unit (1) to receive motion control commands. The power input pin of the first motor drive chip is electrically connected to the motor drive power output terminal provided by the power supply and energy management unit (4). The bridge output pin of the first motor drive chip is connected to the two winding terminals of the DC motor in the micro-injection pump to drive its piston to reciprocate to inject or extract liquid tracer. The external trigger pin of the high-voltage pulse generating chip is connected to the main control and computing unit (1). The general-purpose input / output port of the computing unit (1) is electrically connected to receive the discharge trigger signal. The high-voltage power supply pin of the high-voltage pulse generator chip is connected to the high-voltage DC power supply through the current-limiting resistor. The ground pin of the high-voltage pulse generator chip is grounded. The high-voltage output pin of the high-voltage pulse generator chip is connected to the high-voltage electrode of the microcavity electroporation release device. The ground electrode of the microcavity electroporation release device is grounded. The internal cavity of the microcavity electroporation release device is used to contain capsules encapsulated with engineered microorganisms and generate a strong electric field when receiving a high-voltage pulse to break down the capsule membrane to achieve controllable release. The fluid outlet of the micro-injection pump is used to connect to the injection nozzle located in the water body through a pipeline. The fault status output pin of the first motor drive chip is electrically connected to the interrupt input pin of the main control and computing unit (1) for feedback of the drive status.

5. A mobile water quality monitoring system according to claim 1, characterized in that, The active disturbance generation unit (2) further includes a synthetic turbulence field module (203), which includes a three-phase full-bridge driver chip, first to sixth power MOSFETs, first to third bootstrap diodes, first to third bootstrap capacitors, a three-phase magnetohydrodynamic thruster, a current detection chip, and a thermistor. The three high-side logic input pins and three low-side logic input pins of the three-phase full-bridge driver chip are electrically connected to the pulse width modulation signal output port of the main control and computing unit (1) to receive six independent control waveforms. The three high-side gate drive pins of the three-phase full-bridge driver chip are connected to the gates of the first to third power MOSFETs through first to third gate resistors. The three low-side gate drive pins of the three-phase full-bridge driver chip are connected to the gates of the fourth to sixth power MOSFETs through fourth to sixth gate resistors. The anodes of the first to third bootstrap diodes are connected to the driving power supply, and their cathodes are connected to the first to third high-side floating power supply pins of the three-phase full-bridge driver chip. One end of the first to third bootstrap capacitors is connected to the first to third high-side floating power supply pins of the three-phase full-bridge driver chip. The other ends of the first to third bootstrap capacitors are respectively connected to the first to third high-side reference voltage pins of the three-phase full-bridge drive chip. The first to third high-side reference voltage pins are respectively electrically connected to the sources of the fourth to sixth power MOSFETs. The drains of the first to third power MOSFETs are connected to the high-voltage DC bus. The sources of the first to third power MOSFETs are respectively used as three-phase output terminals. The drains of the fourth to sixth power MOSFETs are respectively electrically connected to the three-phase output terminals. The sources of the fourth to sixth power MOSFETs are connected to the current detection pin of the current detection chip and grounded. The three-phase output terminals are respectively connected to the three coil input terminals of the three-phase magnetohydrodynamic thruster. The voltage output pin of the current detection chip is electrically connected to the analog-to-digital converter input channel of the main control and computing unit (1) for feedback of real-time phase current. The thermistor is mounted on the surface of the coil of the three-phase magnetohydrodynamic thruster. The two ends of the thermistor are connected to another analog-to-digital converter input channel of the main control and computing unit (1) for monitoring the coil temperature. The fault feedback pin of the three-phase full-bridge drive chip is electrically connected to the interrupt input pin of the main control and computing unit (1).

6. The mobile water quality monitoring system according to claim 1, characterized in that, The multimodal signal acquisition and processing unit (3) includes a detection module (301), which includes a single-photon avalanche diode detector array, a time-to-digital converter, a first level conversion chip, and a temperature stabilizer. The single-photon avalanche diode detector array includes sixteen independent detection pixel units. The avalanche signal output pin of each detection pixel unit is connected to the sixteen stop signal input channel pins of the time-to-digital converter. The start signal input channel pin of the time-to-digital converter is electrically connected to the synchronization signal output pin of the laser diode driver chip in the active disturbance generation unit (2) to receive the excitation event start signal. The serial peripheral interface clock pin, serial peripheral interface master input slave output pin, and serial peripheral interface serial peripheral interface master input slave output pin of the time-to-digital converter are connected to the serial peripheral interface master input slave output pin and the serial peripheral interface master input slave output pin. The serial peripheral interface host output slave input pin is electrically connected to the serial peripheral interface host port of the main control and computing unit (1) through the first level conversion chip for configuration and data reading. The interrupt request output pin of the time-to-digital converter is electrically connected to the external interrupt input pin of the main control and computing unit (1) to indicate that the data is ready. The temperature stabilizer is mounted on the package shell of the single-photon avalanche diode detector array. The control signal input terminal of the temperature stabilizer is electrically connected to the general-purpose input / output port of the main control and computing unit (1) to receive temperature control commands to maintain the stable operating temperature of the single-photon avalanche diode detector array. The reference clock input pin of the time-to-digital converter is electrically connected to the high-frequency reference clock signal provided by the communication and clock synchronization unit (5).

7. The mobile water quality monitoring system according to claim 1, characterized in that, The multimodal signal acquisition and processing unit (3) further includes a flow field measurement module (302), which includes an image sensor chip, a field-programmable gate array (FPGA) chip, a first temperature sensor, and a first memory. The pixel data output pin of the image sensor chip is connected to the mobile industrial processor interface receiver pin of the FPGA chip through a mobile industrial processor interface channel. The synchronization signal input pin of the image sensor chip is electrically connected to the synchronization signal output pin of the three-phase full-bridge drive chip in the active disturbance generation unit (2) to receive the flow field generation trigger signal. The control interface of the image sensor chip is electrically connected to the integrated circuit bus master controller pin of the FPGA chip to receive configuration commands. The chip is equipped with an image preprocessing accelerator intellectual property core and a particle image velocimetry calculation intellectual property core. The high-speed transceiver pin of the field-programmable gate array chip is electrically connected to the high-speed serial expansion interface of the main control and computing unit (1) through a serial deserializer link to transmit the processed flow field vector data. The general-purpose input / output port of the field-programmable gate array chip is electrically connected to the control and data bus of the first memory for buffering image frames. The first temperature sensor is mounted on the package surface of the image sensor chip. The analog signal output pin of the first temperature sensor is connected to the analog-to-digital converter input channel of the main control and computing unit (1). The global clock input pin of the field-programmable gate array chip is electrically connected to the low jitter differential clock signal provided by the communication and clock synchronization unit (5).

8. The mobile water quality monitoring system according to claim 1, characterized in that, The multimodal signal acquisition and processing unit (3) further includes a spectral detection module (303), which includes a Raman laser driver, a spectrometer charge-coupled device (CCD), a dedicated spectral processing chip, a high-resolution analog-to-digital converter (ADC), a thermoelectric cooler, and a second temperature sensor. The analog dimming control pin of the Raman laser driver is electrically connected to the output pin of the ADC of the main control and computing unit (1) to receive the light intensity control voltage. The laser enable pin of the Raman laser driver is electrically connected to the general-purpose input / output port of the main control and computing unit (1). The synchronous output pin of the Raman laser driver is electrically connected to the external sampling trigger pin of the high-resolution ADC. The photosensitive surface of the CCD receives the Raman scattered light signal from the water body. The analog video signal output pin of the CCD is connected to the analog signal input positive pin and the analog signal input negative pin of the high-resolution ADC. The serial port of the high-resolution ADC... The slave port of the peripheral interface is electrically connected to the host port of the serial peripheral interface of the dedicated spectral processing chip to transmit digitized spectral data. The dedicated spectral processing chip integrates hardware acceleration units for spectral accumulation and averaging, background subtraction and baseline correction. Its high-speed serial data output pin is connected to the high-speed serial expansion interface of the main control and computing unit (1) through the physical layer of the serial deserializer. The cold end of the thermoelectric cooler is mounted on the back of the charge-coupled device package of the spectrometer. Its drive signal input terminal is electrically connected to the pulse width modulation output pin of the dedicated spectral processing chip. The second temperature sensor is mounted on the surface of the charge-coupled device package of the spectrometer. The output pin of the second temperature sensor is connected to the analog-to-digital converter input channel of the dedicated spectral processing chip. The over-temperature alarm pin of the dedicated spectral processing chip is electrically connected to the interrupt input pin of the main control and computing unit (1). The reference clock input pin of the Raman laser driver is electrically connected to the low phase noise reference clock signal provided by the communication and clock synchronization unit (5).

Images