Marine in-situ test probe and test method based on multi-modal data fusion

Abstract

The invention discloses an ocean in-situ test probe and a test method based on multi-modal data fusion, and relates to the technical field of ocean exploration equipment. The probe comprises a conical probe head and a probe body; a laser source, a Raman spectrometer, a gamma energy disperse spectroscopy, a temperature and pressure measuring unit and an inertial measurement sensor are arranged in the probe body, and the probe body can collect Raman scattering signals to generate a Raman spectrum, receive gamma rays to generate a gamma energy spectrum, measure environment temperature and pressure and measure the motion state of the probe; a pore water pressure sensor at the bottom of the probe body measures pore water pressure; a penetration resistance sensor at the cone top of the conical probe is used for measuring the penetration resistance of the probe; the data processing module is used for calculating the baseline drift distance of the gamma energy disperse spectroscopy according to the environment temperature, the pressure and the probe motion state, correcting the gamma energy spectrum baseline and calculating the pollution risk index according to the penetration resistance, the pore water pressure, the Raman spectrum and the gamma energy spectrum.

Claims

1. Marine in-situ testing probe based on multimodal data fusion, characterized by: The probe includes a probe body and a conical probe (18); The probe body is provided with a laser source (19), a Raman spectrometer (9), a gamma spectrometer (10), a temperature and pressure measurement unit (12), and an inertial measurement sensor; the laser source (19) emits laser light and irradiates the sediment through a first window (22) on the side wall of the probe body to generate a Raman scattering signal; the Raman spectrometer (9) is used to collect the Raman scattering signal and generate a Raman spectrum; the gamma spectrometer (10) is used to receive gamma rays and generate a gamma spectrum; the temperature and pressure measurement unit (12) is used to measure ambient temperature and pressure; and the inertial measurement sensor is used to measure the motion state of the probe; The bottom of the probe body is connected to the cone bottom of the conical probe (18) through a pore pressure filter ring (16), and a pore water pressure sensor (15) is provided in the pore pressure filter ring (16) for measuring the pore water pressure; A penetration resistance sensor (17) is provided at the top of the cone probe (18) for measuring the penetration resistance of the probe; The probe body is also provided with a data processing module (21), which is respectively connected to the Raman spectrometer (9), the gamma spectrometer (10), the pore water pressure sensor (15), the penetration resistance sensor (17), the temperature and pressure measurement unit (12) and the inertial measurement sensor for data processing.

2. The marine in-situ testing probe based on multimodal data fusion according to claim 1, characterized in that: The probe body is provided with a waterproof cavity (6) therein, and the electrical components in the probe body are all located in the waterproof cavity (6). The bottom of the waterproof cavity (6) faces the conical probe (18), that is, faces the direction of the sediment. A second window (23) is provided at the bottom of the waterproof cavity (6), and the gamma spectrometer (10) receives gamma rays through the second window (23). A radiation shielding layer (8) is provided on the inner surface of the waterproof cavity (6), and the thickness of the radiation shielding layer (8) gradually increases from the top to the bottom of the waterproof cavity (6). The radiation shielding layer (8) is not provided at the second window (23), so as to reduce the interference of gamma radiation in the external environment on the gamma spectrometer (10).

3. The marine in-situ testing probe based on multimodal data fusion according to claim 1, characterized in that: The data processing module (21) controls the sampling frequencies of the Raman spectrometer (9) and the gamma spectrometer (10) according to the changes in the penetration resistance and the pore water pressure.

4. The marine in-situ testing probe based on multimodal data fusion according to claim 1, characterized in that: The probe body is a hollow cylindrical hydrophobic tube (13), and the outer wall surface of the hydrophobic tube (13) is sprayed with titanium dioxide nanoparticles, which activate the photocatalytic reaction through frictional electricity penetration, thereby decomposing the biofilm.

5. The marine in-situ testing probe based on multimodal data fusion according to claim 1, characterized in that: A side wall friction sensor (14) is provided on the outer side wall of the probe body for measuring the side wall friction of the probe.

6. The marine in-situ testing probe based on multimodal data fusion according to claim 1, characterized in that: The probe body is further provided with a power management module (5), which includes a main control unit and a battery unit. The main control unit is responsible for power regulation and energy consumption control, and the battery unit is responsible for power supply. The probe body is also provided with a memory (11), which is respectively connected to the Raman spectrometer (9), the gamma spectrometer (10), the pore water pressure sensor (15), the penetration resistance sensor (17), the temperature and pressure measurement unit (12), the inertial measurement sensor and the data processing module (21) for data storage.

7. The marine in-situ testing probe based on multimodal data fusion according to claim 1, characterized in that: The probe body is further provided with a signal amplifier (7), a preamplifier (4) and an analog-to-digital converter (3) connected in sequence. The original signals output by the Raman spectrometer (9) and the gamma spectrometer (10) are respectively preliminarily amplified by the signal amplifier (7), further amplified by the preamplifier (4), and converted by the analog-to-digital converter (3), and then transmitted to the external device via the cable (1); The pore water pressure sensor (15), the penetration resistance sensor (17), the temperature and pressure measurement unit (12), the inertial measurement sensor, and the data processing module (21) are also connected to external devices via the cable (1).

8. The marine in-situ testing method based on multimodal data fusion is characterized by: The method is applicable to the marine in-situ test probe based on multimodal data fusion according to any one of claims 1 to 7 above, and is specifically as follows: S1, when the conical probe (18) of the probe penetrates the sediment, the laser source (19) emits laser light and irradiates the sediment to generate Raman scattering signals, the Raman spectrometer (9) receives the Raman scattering signals and generates a Raman spectrum; the detector of the gamma spectrometer (10) receives gamma rays and generates a gamma spectrum; the pore water pressure sensor (15) measures the pore water pressure; the penetration resistance sensor (17) measures the penetration resistance of the probe; the temperature and pressure measurement unit (12) measures the ambient temperature and pressure; the inertial measurement sensor measures the motion state of the probe; and the various types of data collected are sent to the data processing module (21); S2, the data processing module (21) calculates the baseline drift ΔE of the gamma spectrometer (10) according to the ambient temperature, pressure and the motion state of the probe, thereby correcting the gamma spectrum baseline; S3, the data processing module (21) uses a pollution risk assessment algorithm to conduct a comprehensive analysis of the penetration resistance, pore water pressure, Raman spectrum characteristic peaks and gamma energy spectrum peaks to obtain the pollution risk index PRI.

9. The marine in-situ testing method based on multimodal data fusion according to claim 8, characterized in that: In step S3, the data processing module (21) uses the random forest algorithm to calculate the PRI value. The specific calculation method is as follows: Where X is the characteristic vector composed of pore water pressure, penetration resistance, Raman spectrum characteristic peak intensity, and γ energy spectrum peak; N is the number of decision trees; h i (X) represents the output result of the i-th decision tree for the input feature vector X.

10. The marine in-situ testing method based on multimodal data fusion according to claim 8, characterized in that: The calculation method of the baseline drift ΔE is: ΔE = k1ΔT + k202ΔP - k3v, where ΔT is the temperature change rate, ΔP is the pressure change rate, v is the penetration velocity, and k1, k2, and k3 are all coefficients.

Images