Temperature-salt sensor and real-time error compensation method thereof

Abstract

The invention relates to an ocean sensor technology, in particular to a temperature and salt sensor and a real-time error compensation method thereof, and the compensation method comprises the following steps: synchronously collecting temperature data, conductivity data, water depth pressure data and auxiliary environment parameters of seawater; calculating the movement speed of the equipment through the pressure variation of the adjacent sampling points; decomposing the error into a horizon deviation error, an environment interference error and an equipment movement speed related error; on the basis of an error decomposition result, calculating compensation parameters through a multivariate error model, and performing two-dimensional compensation on the temperature data and the conductivity data according to the compensation parameters, including time matching compensation based on sensor response time difference and correction for environmental parameter influence; based on the corrected temperature and conductivity data, the salinity is calculated through a standard seawater state equation according to the ITS-90 temperature standard and the PSS-78 salt standard, and therefore the measurement precision and stability are remarkably improved.

Description

Technical Field

[0001] This invention relates to marine sensor technology, specifically a temperature and salinity sensor and its real-time error compensation method. Background Technology

[0002] Currently used temperature, salinity, and thermal (TCT) sensors generally employ simple calibration models. These models typically suffer from static limitations, making them ill-suited to the dynamic complexity of the marine environment and the variations in sensor response characteristics. Specifically: 1) Temperature-salinity coupling error: Changes in temperature and salinity can affect each other, easily leading to systematic deviations in sensor data; 2) Response time mismatch: The pressure sensor responds the fastest and the temperature sensor responds the slowest. When the sensors move rapidly with the mobile equipment or the profile measurement equipment (such as when the lowering speed reaches 3-4 m / s), the measurement data of the three correspond to different seawater layers, causing "salinity spike" errors in the salinity calculation. 3) Environmental interference superposition: Pressure fluctuations, water flow impacts and the adhesion of floating matter further reduce the stability of the sensor, while existing compensation methods mostly rely on hardware design or static parameter correction, and cannot be dynamically adjusted in combination with the movement state of the equipment.

[0003] In addition, existing compensation methods for salinity peaks (such as the GM method and the Grose method) are only suitable for scenarios where the equipment is deployed slowly. In rapid dynamic measurements, overcompensation is likely to occur, which further affects the accuracy of the data. Summary of the Invention

[0004] The present invention aims to provide a temperature and salinity sensor and its real-time error compensation method to overcome the measurement errors of traditional sensors in different marine environments, thereby improving the measurement accuracy and stability of salinity and temperature.

[0005] To achieve the above objectives, the technical solution specifically adopted by the present invention is as follows: A method for compensating for real-time errors in a temperature-salt sensor includes the following steps: Step 1: Multi-dimensional data collection The temperature data T, conductivity data C, and water depth and pressure data P of seawater are collected simultaneously by temperature sensors, conductivity sensors, and pressure sensors. At the same time, auxiliary environmental parameters, including water flow velocity v, water turbidity, and suspended matter concentration p, are collected by multi-parameter probes or external sensors. Step 2: Pressure Correction First, the collected water depth and pressure data P is processed by moving average filtering to eliminate pressure anomalies caused by hull undulation or water flow impact; then, the pressure change between adjacent sampling points is calculated. ( ,in The filtered pressure data (in Pa) for the (n+1)th and nth sampling points are respectively, and based on this pressure change... The speed of the computing device ;in, The velocity of the equipment (in seawater) (such as the lowering and rising speed), in m / s, is used to quantify the real-time motion status of the equipment. The system sampling interval is determined by the sampling frequency of 25Hz. The unit is This speed data will serve as a core parameter, providing crucial input for subsequent error decomposition (quantification of speed-related errors) and two-dimensional compensation (dynamic adjustment of lag compensation).

[0006] Step 3: Error Decomposition Error analysis is performed using an algorithmic compensation model, decomposing measurement errors into three core types of errors: (1) Layer deviation error caused by sensor response time mismatch, which is based on the pressure sensor response time. For reference zero point, including temperature sensor hysteresis time hysteresis time of conductivity sensor ; (2) Environmental interference errors, mainly including conductivity errors caused by temperature drift, contact resistance changes caused by planktonic obstruction, and conductivity cell constant shifts caused by seawater pressure changes; (3) The error related to the speed of equipment movement is that the faster the speed, the greater the layer deviation error. Based on the error decomposition results, the error correction parameters are calculated using a multivariate error model, which can be expressed as: ,in These are the response times of temperature, conductivity, and pressure sensors, respectively. This is historical temperature and salinity profile data.

[0007] Step 4: Two-dimensional compensation Based on the compensation parameters output by the multivariate error model, the temperature data T and conductivity data C are compensated in real time in two dimensions, including: (1) Response time matching correction: Pressure sensor response time Using the zero point as a reference, and combining the equipment's movement speed v to dynamically adjust the hysteresis compensation of temperature and conductivity data, the correction formula is as follows: ; ; in, The dynamic offset is based on the sampling interval. Speed ​​correction factor , satisfy , satisfy By traversing the response time range of the temperature sensor (100-130 ms) and the response time range of the conductivity sensor (25-35 ms), the parameter combination with the minimum salinity peak and the optimal vertical gradient was selected as... , The optimal value; (2) Environmental noise correction: A 5-point moving average filter was used to remove electrode disturbance noise, and historical temperature and salinity profile data were fused using a Kalman filter. Optimize environmental noise weighting coefficient The temperature and conductivity data after response time matching are then corrected a second time. Step 5: Salinity Calculation: Based on the corrected temperature T' and conductivity C' data, using the PSS-78 salinity standard and corrected for temperature using the ITS-90 temperature scale, the salinity S was calculated using the standard seawater state equation (TEOS-10 compatible with PSS-78). The salinity calculation formula is as follows: ; in, , It is the ratio of the conductivity of seawater at 15℃ and standard atmospheric pressure to the conductivity of standard KCl solution.

[0008] Step 6: Data Interpolation and Output The compensated salinity data is subjected to vertical linear interpolation to compress the data into standardized data at a preset resolution, including but not limited to 1m resolution. The data is transmitted to a remote monitoring system or data platform via wireless (LoRa / NB-IoT) or wired (RS485 / CAN) methods. The output data includes compensated temperature T', conductivity C', salinity S, and original parameters (collection time, water depth, pressure, and movement speed). It supports data integration from devices such as marine buoys and mobile navigation systems.

[0009] Furthermore, in step 1, the temperature sensor has a response time of 100±5ms, a range of -5 to 30℃, and an accuracy of ±0.003℃; the conductivity sensor has a response time of 25±3ms, a range of 0 to 70mS / cm, and an accuracy of ±0.005mS / cm; the pressure sensor has a response time of 10±2ms, a range of 0 to 1000m, and an accuracy of ±0.05%FS; the system sampling frequency is 25Hz, and the sampling interval Δt=40ms, ensuring the time synchronization of data acquisition.

[0010] The present invention also provides a temperature-salt sensor, which uses the above-described compensation method to achieve real-time error compensation.

[0011] The present invention provides a temperature-salt sensor, comprising: Temperature sensor used to measure seawater temperature in real time.

[0012] Conductivity sensor: used for real-time measurement of seawater conductivity; Pressure sensor: used to measure water depth and pressure data in real time; Signal processing module: It has a built-in 16-bit analog-to-digital converter (ADC), low-noise amplifier circuit and 32-bit microprocessor (MCU), supports 25Hz high-frequency data processing, and has a built-in synchronous trigger unit to ensure the data acquisition time alignment of temperature sensor, conductivity sensor and pressure sensor (sampling interval error ≤1ms), and is used for sensor signal preprocessing, moving average filtering and preliminary compensation. The data transmission module adopts RS485, CAN, LoRa or NB-IoT communication modules, supports real-time data transmission at a high frequency of 25Hz, and has a transmission delay of ≤100ms. It is used to output the final compensated temperature and salinity data to the host computer or remote monitoring platform.

[0013] The temperature sensor and conductivity sensor are fixed at the front end of the sensor probe, respectively; the pressure sensor is integrated at the rear end of the sensor probe, and the sensor probe as a whole adopts a streamlined design.

[0014] This invention provides a method for compensating for real-time errors in a temperature, salinity, and thermal (TST) sensor. This method significantly improves the measurement accuracy and stability of the TST sensor and is applicable to dynamic monitoring scenarios such as marine monitoring, climate change research, marine resource exploration, and mobile oceanographic profiling and vertical profile observation. It is compatible with various marine survey equipment, including MVP (Mobile Multi-Parameter Oceanographic Measurement System) and conventional CTD (Conductivity, Tolerance, and Difference) systems, and has the following characteristics and beneficial effects: (1) High-precision correction: For dynamic measurement scenarios (equipment movement speed 0.5-4m / s), it effectively eliminates salinity spike errors. Compared with the industry standard SBE-9 CTD data, the salinity cross-correlation coefficient after compensation is ≥0.91, the maximum salinity difference is ≤0.2PSU, and the error is reduced by more than 80% compared with the uncompensated version. In static monitoring scenarios, the salinity error is reduced to 0.02-0.05PSU, which is better than the traditional method (0.15-0.30PSU).

[0015] (2) Strong stability guarantee: The algorithm can still work stably in complex marine environments with current velocities of 1.2-4 m / s through dynamic velocity correction coefficient and parameter optimization strategy, without overcompensation, thus solving the adaptation defects of existing GM method and Grose method in rapid dynamic measurement.

[0016] (3) Full-scene adaptability: It is compatible with both fixed monitoring (such as ocean buoys) and mobile / profiling dynamic monitoring (such as MVP system). The horizontal spatial resolution can reach 1.8km, which can clearly present marine micro-phenomena such as thermo-salinity strata and cold water fronts, and adapt to the measurement needs under different water layers and different ocean current conditions.

[0017] (4) Improved data accuracy: After compensation, the false low-salinity areas at the thermocline completely disappeared, and the isohaline distribution more closely matches the actual marine environment. Attached Figure Description

[0018] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 This is a schematic diagram of the structure of a temperature and salinity sensor according to an embodiment of the present invention; In the diagram: 1-Temperature sensor; 2-Conductivity sensor; 3-Signal processing module; 4-Data transmission module; 5-Pressure sensor.

[0019] Figure 2 This is a flowchart illustrating a method for compensating for real-time errors in a temperature-salt sensor according to an embodiment of the present invention. Detailed Implementation

[0020] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0021] The first aspect of the present invention is to provide a temperature-salt sensor, such as... Figure 1 As shown, the temperature-salt sensor mainly includes: Temperature sensor 1: It adopts a platinum resistance thermometer (PT100 / PT1000) or a thermistor, which is encapsulated in a corrosion-resistant probe. The response time is 100±5ms, the range is -5~30℃, and the accuracy is ±0.003℃. It is used for real-time measurement of seawater temperature.

[0022] Conductivity sensor 2: It adopts a four-electrode or inductive conductivity structure, with a response time of 25±3ms, a range of 0~70mS / cm, and an accuracy of ±0.005mS / cm. It is used to measure the conductivity of seawater in real time and avoids errors caused by polarization effects.

[0023] Pressure sensor 5: It adopts a piezoresistive pressure sensor with a response time of 10±2ms, a range of 0~1000m, an accuracy of ±0.05%FS, an output signal compatible with 16-bit ADC, supports 25Hz sampling, is easy to integrate, and is used for real-time measurement of water depth and pressure data. Signal processing module 3: It has a built-in 16-bit analog-to-digital converter (ADC), low-noise amplifier circuit and 32-bit microprocessor (MCU), supports 25Hz high-frequency data processing, and has a built-in synchronous trigger unit to ensure that the data acquisition time of temperature sensor, conductivity sensor and pressure sensor is aligned (sampling interval error ≤1ms), and is used for sensor signal preprocessing, moving average filtering and preliminary compensation.

[0024] Data transmission module 4: It adopts RS485, CAN, LoRa or NB-IoT communication modules, supports real-time transmission of high-frequency data at 25Hz, and has a transmission delay of ≤100ms. It meets the data transmission requirements of high-level resolution observation of mobile systems and outputs the finally compensated temperature and salinity data to the host computer or remote monitoring platform.

[0025] To adapt to mobile / profile measurement equipment, the sensor probe adopts a streamlined design to reduce the impact of water flow on the measurement. Temperature sensor 1 and conductivity sensor 2 are fixed at the front end of the sensor probe, with a distance of 1-5cm between them; pressure sensor 5 is integrated at the rear end of the sensor probe.

[0026] The temperature and salinity sensor collects data according to the following procedure: (1) Temperature acquisition: The temperature sensor acquires temperature data every 5 degrees Celsius, and transmits it to the MCU after analog-to-digital conversion.

[0027] (2) Conductivity acquisition: The conductivity sensor operates at a frequency of 1Hz~10Hz to acquire changes in electrode voltage and thus calculate the conductivity of seawater.

[0028] (3) Water depth pressure acquisition: The piezoresistive pressure sensor is integrated at the rear end of the sensor probe. It senses pressure changes by directly contacting seawater. Its core sensitive element (piezoresistive bridge) converts the pressure signal into a weak voltage signal. After the voltage signal is initially amplified by the built-in signal conditioning circuit of the sensor, it is transmitted to the low-noise amplification circuit of the signal processing module for further gain adjustment. Then, the analog signal is converted into a digital signal by the 16-bit analog-to-digital converter chip (ADC). The synchronous triggering unit ensures that the pressure data acquisition is time-aligned with the temperature and conductivity data (sampling interval error ≤ 1ms). The raw water depth pressure data P is continuously acquired at a sampling frequency of 25Hz (Δt = 40ms) to provide basic data for subsequent pressure correction and equipment movement speed calculation.

[0029] (4) Synchronization Processing: The signal processing module uses a built-in synchronization trigger unit to time-align temperature data, conductivity data, and water depth and pressure data (sampling interval error ≤ 1ms) to ensure that the three correspond to the same seawater layer. Subsequently, the aligned temperature and conductivity values ​​are converted into Practical Salinity (PSU) according to marine survey standards and PSS-78 salinity scale and ITS-90 temperature scale, combined with environmental parameters corrected by water depth and pressure data. At the same time, the original synchronized water depth and pressure data and the calculated equipment movement speed are retained to provide a time-synchronized basic dataset for subsequent pressure correction and error compensation.

[0030] The second aspect of this invention is to provide a real-time error compensation algorithm for temperature and salinity sensors that integrates "sensor response time matching + environmental parameter adaptation." This compensation algorithm uses the pressure sensor response time as a benchmark, dynamically adjusts the hysteresis compensation of temperature and conductivity data based on the real-time movement speed of the device (calculated based on pressure changes), and simultaneously couples environmental parameters such as water flow and suspended matter to correct noise errors, thus solving the salinity spike problem in dynamic measurement scenarios. Specifically: like Figure 2 As shown, this compensation method includes the following core steps: Step 1: Data Collection Collect seawater temperature data (T), electrical conductivity data (C), water depth and pressure data (P), and auxiliary parameters from the environmental model, such as: water flow velocity v Water turbidity and suspended solids concentration p Historical temperature and salinity profile data

[0031] The seawater temperature data (T), conductivity data (C), and water depth / pressure data (P) are acquired synchronously by temperature, conductivity, and pressure sensors. The system sampling frequency is 25Hz, and the sampling interval Δt = 40ms. The synchronization trigger unit of the signal processing module ensures data acquisition time alignment. This can be provided by a multi-parameter probe or an external sensor; specifically: The water flow velocity v is provided by an electromagnetic flow meter or an ultrasonic Doppler velocimeter. Measurement range: 0~5m / s; measurement accuracy: ±1%FS; response time: ≤50ms; operating frequency: 10~100Hz; output signal: 4-20mA or pulse signal. The water turbidity and suspended solids concentration (p) are provided by a diffuse light turbidity sensor. Measurement range: 0~1000 NTU; Measurement accuracy: ±2%FS; Light source type: Output signal: 4-20mA or RS485 digital signal. Historical temperature and salinity profile data Data is sourced from the CTD historical database or ocean observation network. Data format: depth-temperature-salinity three-dimensional profile data; update frequency: real-time or daily; spatial resolution: horizontal resolution 1km×1km, vertical resolution 1m; temporal resolution: minute to hour; storage format: NetCDF or CSV standard format.

[0032] Step 2: Pressure Correction 2.1 The water depth and pressure data P are processed by moving average filtering to eliminate outliers caused by hull undulation or water flow impact; 2.2 Calculate the pressure change between adjacent sampling points And based on this pressure change The speed of the computing device ; Step 3: Error decomposition: 3.1 The measurement error is decomposed into three core types of errors through the error analysis module (with built-in algorithm compensation model) embedded in the MCU: Response time mismatch error: temperature Electrical conductivity ,pressure Layer deviation caused by sensor response time mismatch, Used as a reference zero point; Environmental interference error: conductivity error caused by temperature drift, mainly including electrode disturbance noise caused by ocean current impact, contact resistance change caused by planktonic obstruction, and conductivity cell constant shift caused by seawater pressure change; Speed-related error: The greater the equipment movement speed v, the more significant the layer deviation error; 3.2 Based on the error decomposition results, the error correction parameters are calculated using a multivariate error model. The error model can be expressed as: ,in These are the response times of temperature, conductivity, and pressure sensors, respectively. This is historical temperature and salinity profile data.

[0033] It is worth noting that the embodiments of the present invention use a multivariate regression model, but a Kalman filter model, a BP neural network, or an LSTM ocean time series prediction model can also be selected to adapt to different deployment environments.

[0034] Step 4: Application of the Two-Dimensional Compensation Algorithm The algorithm module outputs compensation parameters based on the multivariate error model, including the temperature compensation coefficient. Conductivity compensation coefficient Environmental noise weighting coefficient Speed ​​correction coefficient The original data was corrected using a two-dimensional approach: "response time matching + environmental noise correction". (1) Response time matching correction: Pressure sensor response time Using the zero point as a reference, and combining the equipment's movement speed v to dynamically adjust the hysteresis compensation of temperature and conductivity data, the correction formula is as follows: ; ; Where k = 0~2 is the dynamic offset based on the sampling interval. Speed ​​correction factor , satisfy , satisfy By traversing the response time range of the temperature sensor (100-130 ms) and the response time range of the conductivity sensor (25-35 ms), the parameter combination with the minimum salinity peak and the optimal vertical gradient was selected as... , The optimal value; (2) Environmental noise correction: A 5-point moving average filter was used to remove electrode disturbance noise, and historical temperature and salinity profile data were fused using a Kalman filter. Optimize environmental noise weighting coefficient A secondary correction is performed on the temperature T' and conductivity C' after the response time is matched; Step 5: Salinity Calculation: Based on the corrected temperature T' and conductivity C' data, using the PSS-78 salinity standard and corrected for temperature using the ITS-90 temperature scale, the salinity S was calculated using the standard seawater state equation (TEOS-10 compatible with PSS-78). The salinity calculation formula is as follows: ; in, , It is the ratio of the conductivity of seawater at 15℃ and standard atmospheric pressure to the conductivity of standard KCl solution.

[0035] Step 6: Data Interpolation and Output The compensated salinity data is subjected to vertical linear interpolation at a resolution of 1m, and compressed into standardized data. The compensated temperature T', conductivity C', and salinity S are output to a host computer or remote platform via a data transmission module, supporting: real-time serial printing (RS485), wireless upload (LoRa / NB-IoT), and data integration with marine buoys or navigation systems. Example of output data format: T=18.52°C; C = 4.221 mS / cm; S = 33.87 PSU; Time=2025-01-01 10:25:22. To verify the effectiveness of the system of the present invention, two types of application cases were set up for comparative experiments: Example 1: Static Fixed Monitoring The sensor was deployed at a depth of 6000m in a nearshore area and operated continuously for 30 days. The effects of the compensation algorithm of this invention were compared with those of the traditional uncompensated method. Traditional methods have a salinity error of approximately 0.15–0.30 PSU; After the compensation algorithm of this invention, the salinity error is reduced to 0.02–0.05 PSU; It can maintain high stability even when the flow rate is greater than 1.2 m / s, with data fluctuation amplitude ≤ 0.03 PSU.

[0036] Example 2: Dynamic mobile profiling measurement The sensor was integrated into the MVP-200 mobile system, and observations were conducted in the 35°N section (120.5°~123.5°E). The device was lowered at a speed of 3-4 m / s, and a total of 385 vertical profile data points were acquired. The data were compared with the industry standard SBE-9 CTD data. Without compensation, a distinct salinity peak appears at the thermocline, with a maximum difference of 1.2 PSU compared to the SBE-9 type CTD data; After compensation by this invention, the salinity peak completely disappears, the maximum difference with SBE-9 type CTD data is ≤0.2PSU, the cross-correlation coefficient reaches 0.917, and the error is reduced by 80%; Compared to existing GM and Grose methods, this invention does not exhibit overcompensation, and the clarity of the cold water mass front structure near 123°E is superior to that of SBE-9 type CTD data (with denser isosal lines and richer details). The horizontal spatial resolution reaches 1.8km, meeting the requirements for high-resolution observation of marine phenomena.

[0037] Based on Examples 1 and 2, it can be demonstrated that the compensation algorithm of the present invention can significantly improve measurement accuracy and anti-interference ability in both static and dynamic scenarios, and its adaptability and practicality are superior to the prior art.

[0038] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for compensating for real-time errors in a temperature-salinity sensor, characterized in that: Includes the following steps: Step 1: Multi-dimensional data collection Simultaneously collect seawater temperature data (T), electrical conductivity data (C), water depth and pressure data (P), and auxiliary environmental parameters; Step 2: Pressure Correction After filtering the collected water depth and pressure data, the movement speed of the equipment is calculated based on the pressure change at adjacent sampling points. Step 3: Error Decomposition The measurement error is decomposed into layer deviation error caused by sensor response time mismatch, environmental interference error, and equipment movement speed-related error. Based on this error decomposition result, compensation parameters are calculated through a multivariate error model. The error model can be expressed as: ； in, , , These are the response times of temperature, conductivity, and pressure sensors, respectively. Historical temperature and salinity profile data; Step 4: Two-dimensional compensation Based on the compensation parameters output by the multivariate error model, the temperature data T and conductivity data C are compensated in real time in two dimensions, including time matching compensation based on the difference in sensor response time and correction for the influence of environmental parameters. Step 5: Salinity Calculation Based on the corrected temperature T' and conductivity C' data, the PSS-78 salinity standard was adopted, and the temperature was corrected by the ITS-90 temperature standard. The salinity S was calculated using the standard seawater state equation. Step 6: Data Interpolation and Output After interpolation processing, the compensated salinity data is transmitted to a remote monitoring system or data platform.

2. The method for compensating for real-time errors of a temperature-salinity sensor as described in claim 1, characterized in that: In step 1, temperature data T, conductivity data C, and water depth / pressure data P of seawater are simultaneously collected using a temperature sensor, a conductivity sensor, and a pressure sensor. The temperature sensor has a response time of 100±5ms, a range of -5 to 30℃, and an accuracy of ±0.003℃; the conductivity sensor has a response time of 25±3ms, a range of 0 to 70mS / cm, and an accuracy of ±0.005mS / cm; and the pressure sensor has a response time of 10±2ms, a range of 0 to 1000m, and an accuracy of ±0.05%FS. The system sampling frequency is 25Hz, and the sampling interval Δt=40ms to ensure the time synchronization of data acquisition.

3. The method for compensating for real-time error of a temperature-salinity sensor as described in claim 1, characterized in that: In step 1, auxiliary environmental parameters, including water flow velocity v, water turbidity, and suspended matter concentration p, are collected by a multi-parameter probe or external sensor.

4. The method for compensating for real-time error of a temperature-salinity sensor as described in claim 1, characterized in that: In step 4, time-matching compensation based on sensor response time differences is performed: using the pressure sensor response time... Using the zero point as a reference, and combining the equipment's movement speed v to dynamically adjust the hysteresis compensation of temperature and conductivity data, the correction formula is as follows: ； ； in, This is a dynamic offset based on the sampling interval. For speed correction factor, satisfy , satisfy By traversing the response time range of the temperature sensor (100-130 ms) and the response time range of the conductivity sensor (25-35 ms), the parameter combination with the minimum salinity peak and the optimal vertical gradient was selected as... , The optimal value.

5. The method for compensating for real-time error of a temperature-salinity sensor as described in claim 1, characterized in that: In step 4, the correction for the influence of environmental parameters is as follows: a 5-point moving average filter is used to remove electrode disturbance noise, and historical temperature and salinity profile data is fused using Kalman filtering. Optimize environmental noise weighting coefficient The temperature and conductivity data after the response time matching are then corrected a second time.

6. The method for compensating for real-time error of a temperature-salinity sensor as described in claim 1, characterized in that: In step 5, the salinity calculation formula is as follows: ； in, , It is the ratio of the conductivity of seawater at 15℃ and standard atmospheric pressure to the conductivity of standard KCl solution.

7. The method for compensating for real-time error of a temperature-salinity sensor as described in claim 1, characterized in that: In step 6, the data interpolation uses a vertical linear interpolation algorithm to compress the data into standardized data at a preset resolution, which includes, but is not limited to, a 1m resolution.

8. A temperature-salt sensor, characterized in that, The sensor uses the compensation method described in any one of claims 1-7 to compensate for real-time errors.

9. A temperature and salinity sensor as described in claim 8, characterized in that, include: Temperature sensor for real-time measurement of seawater temperature; Conductivity sensor: used for real-time measurement of seawater conductivity; Pressure sensor: used to measure water depth and pressure data in real time; Signal processing module: It has a built-in 16-bit analog-to-digital converter chip, low-noise amplifier circuit and 32-bit microprocessor, supports 25Hz high-frequency data processing, and has a built-in synchronous trigger unit to ensure the data acquisition time alignment of temperature sensor, conductivity sensor and pressure sensor, and is used for sensor signal preprocessing, moving average filtering and preliminary compensation. The data transmission module, using RS485, CAN, LoRa or NB-IoT communication modules, is used to output the final compensated temperature and salinity data to the host computer or remote monitoring platform. The temperature sensor and conductivity sensor are fixed at the front end of the sensor probe, respectively; the pressure sensor is integrated at the rear end of the sensor probe, and the sensor probe as a whole adopts a streamlined design.

Images