A multi-point distributed thermocouple array and real-time temperature field monitoring method

Abstract

The application discloses a kind of multipoint distributed thermocouple array and real-time temperature field monitoring method, it is related to temperature measurement technical field.The array of the present application is integrated micro-thermoelectric junction on substrate with >1000 point cm-2 Density, subarea uses Bi2Te3 / Ag, Ni-Cr-Si / Cu-Ni and W-26%Re / Pt-13%Rh, realize-200°C to 2000°C temperature measurement.Row column strobe-FPGA pipeline architecture completes full array scanning and IDW interpolation within 1ms, and output high-resolution temperature field.Fiber or 1.6GHz backscattering link guarantees >900°C and EMI environment reliable data transmission;Visual interface <10µs triggers alarm or executes closed-loop cooling.It is suitable for aeroengine, high-temperature industry and biomedical precise hyperthermia.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of temperature measurement, in particular to a multi-point distributed thermocouple array and a real-time temperature field monitoring method. BACKGROUND

[0002] The existing multi-point temperature measurement mainly relies on three major routes of single material linear thermocouples, thin film thermocouple arrays and optical fiber distributed temperature sensing. Although single material thermocouples are resistant to high temperature, they can only be sparsely distributed, require complex wiring, and have reduced sensitivity below -200°C; thin film arrays are limited to ≤100°C flexible electronics or ≤950°C noble metal segments, have few points, slow scanning, often use neural network compensation for crosstalk, and have large delay in calculation; optical fiber DTS can continuously measure temperature at long distances, but is restricted by pulse width and scattered signal intensity, with a typical spatial resolution of ≥1m, which cannot meet the sub-millimeter demand of turbine blades or tissue ablation. In addition, existing solutions often require expensive heat-resistant leads or shielding in high-temperature environments of ≥900°C and strong electromagnetic interference, resulting in high overall system cost and large maintenance volume. SUMMARY

[0003] To overcome the defects of the prior art, the present application provides the following technical solution: a multi-point distributed thermocouple array, the array comprising: a multi-material gradient thermocouple array unit forming a matrix of >1000 thermoelectric junctions cm⁻² on a substrate, and using a first thermoelectric material pair covering a temperature range of -200°C to 0°C, a second thermoelectric material pair covering a temperature range of 0°C to 900°C, and a third thermoelectric material pair covering a temperature range of 900°C to 2000°C in different zones to realize full-temperature range measurement from -200°C to 2000°C; a cold-end constant temperature reference zone with a temperature stabilized at a set reference value and electrically connected to all cold ends of the thermocouple array unit through heat-conducting wires; a high-speed row-column selection signal acquisition module including a row-column selection switch matrix, a low-noise amplifier and a high-resolution analog-to-digital converter for the thermocouple array unit, configured to complete potential reading of all nodes of the thermoelectric junction matrix within ≤1ms; a hardware deterministic temperature field reconstruction module including a lookup table linearization circuit and an inverse distance weighted interpolation circuit, configured to output a gridded temperature field matrix immediately after receiving the read data; an optical and / or wireless signal transmission link configured to transmit the read data to a safe area through optical or radio frequency signals in environments where the temperature is greater than 900°C or electromagnetic interference exists; a visualization and alarm interface configured to output an alarm signal when the temperature exceeds the limit and to display the temperature field matrix in real time.

[0004] Preferably, the first thermoelectric material pair is Bi2Te3 / Ag, suitable for -200°C to 0°C; the second thermoelectric material pair is Ni-Cr-Si / Cu-Ni, suitable for 0°C to 900°C; the third thermoelectric material pair is tungsten-26% rhenium / platinum-13% rhodium, suitable for 900°C to 2000°C.

[0005] Preferably, the thermoelectric junction array unit is prepared by a MEMS-sputtering-3D direct writing composite process, and the area of the thermoelectric junction unit is .

[0006] Preferably, the row-column gating switch matrix is a GaN-HEMT board-level array, the driving clock is ≥40MHz, and the crosstalk suppression ratio is ≥80dB.

[0007] Preferably, the interpolation circuit of the hardware deterministic temperature field reconstruction module adopts an FPGA parallel pipeline, achieving a frame rate of ≥1kHz and a same-frame delay of ≤50µs.

[0008] Preferably, the optical signal transmission link is completed through a 1550nm wavelength division multiplexing optical fiber, and the data bandwidth is ≥10Gbps; the wireless link adopts a 1.6GHz backscatter, and the transmission distance is ≤50cm.

[0009] Preferably, the visualization and alarm interface outputs a TTL alarm with a delay of <10µs and sends a Modbus-TCP control word synchronously when the temperature exceeds a set threshold.

[0010] A real-time temperature field monitoring method for a multi-point distributed thermocouple array, characterized by comprising the following steps: S1, deploying a multi-material gradient thermocouple array unit on the surface or inside of an object to be measured and connecting it with a cold-end constant-temperature reference zone; S2, starting a high-speed row-column gating signal acquisition module to read all thermoelectric junction output potentials at a period of ≤1ms; S3, performing linearization conversion and interpolation processing on the read potentials in a hardware deterministic temperature field reconstruction module and outputting a temperature field matrix; S4, sending the temperature field matrix to an external visualization and alarm interface through an optical and / or wireless transmission link; S5, outputting an alarm or control signal when the temperature of any point in the temperature field matrix exceeds a preset threshold.

[0011] Preferably, the linearization conversion in step S3 adopts a fifth-order polynomial lookup table method, and the single-point conversion time is ≤100ns.

[0012] Preferably, the interpolation processing in step S3 adopts an inverse distance weighted interpolation algorithm, the weight exponent is 2, and the interpolation range does not exceed 5×5 adjacent temperature measurement pixels.

[0013] Compared with the prior art, the present invention has the following advantages: (1) The present invention achieves continuous temperature measurement range coverage from -200°C to 2000°C in the same array by using the first, second and third thermoelectric materials, eliminating the complicated process of segmented deployment and switching calibration of traditional multi-probes; it maintains ±0.5% reading error in the ultra-wide temperature range, so that extremely low temperature and extremely high temperature conditions can be presented on a single "temperature map", greatly improving the efficiency of integrated measurement and control; (2) The array of the present invention forms >1000 points on the substrate. Thermoelectric junction matrix, single junction area The spatial resolution is improved to the sub-millimeter level; compared with traditional single-point or centimeter-level infrared temperature measurement, the fineness of thermal gradient imaging is significantly improved, providing a data foundation for fine thermal management; (3) The row and column gating scheme of this invention adopts GaN-HEMT switch matrix and parallel pipeline FPGA, with a complete frame reading time of 51.2µs and a frame rate of 1kHz. Combined with the intrinsic µs thermal inertia brought by the micro-nano thermal capacity structure, it can resolve the transient temperature rise in the process of shock wave, combustion explosion or radio frequency ablation in less than milliseconds, breaking through the limitation of tens of milliseconds of lag of existing thermocouples; (4) The fiber wavelength division multiplexing link and 1.6GHz backscatter link of this invention realize metal leadless transmission for the high temperature area >900°C and the strong electromagnetic interference area, respectively, eliminating the risks of high temperature oxidation, contact resistance drift and EMI noise coupling. Attached Figure Description

[0014] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0015] The technical solution of the present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0016] The real-time temperature field monitoring mechanism provided by the application is based on the Seebeck effect of thermocouples, and realizes real-time and high-precision temperature field reconstruction: direct thermoelectric conversion and differential measurement: each micro-sensing unit is composed of a thermoelectric junction of two different materials, and a millivolt-level potential signal is directly generated by a temperature difference. A common reference node or reference bus is designed in the array: for example, the cold end of all thermocouples is connected to a constant-temperature reference area (or a built-in reference junction in the material). In this way, the output voltage of each sensing point actually represents the temperature difference between the point and the reference. By integrating a micro-constant-temperature tank / reference block at the edge or a specific position of the array, absolute temperature measurement is realized, and the uncertainty of the relative values of the points is avoided. High-frequency multiplexing: in order to obtain the data of all sensing points, a high-speed electronic switch matrix or an integrated multiplexing reading circuit is used. In hardware, the array row and column leads and the thermoelectric junction form an Addressable Matrix architecture, and by means of high-speed scanning row by row and column by column, point-by-point polling reading of thousands of sensors is realized. With the help of a switch array chip or an FPGA, the overall scanning frequency is sub-millisecond. For example, if there are 1000 temperature measurement points, and each point takes 0.1 ms to sample, then Full array scanning can be completed; through multi-channel parallel scanning and hardware pipeline, the total refresh period can be further compressed to milliseconds or even lower. For higher demand, it can also be divided into regional parallel reading (dividing the array into multiple blocks for simultaneous acquisition), so as to realize real-time temperature field update. Low noise high speed amplification: due to the small output voltage of a single micro thermocouple (μV level / °C), the signal conditioning circuit is crucial. Each channel is equipped with a nanometer low noise amplifier and a high speed analog to digital converter, ensuring that it still has sufficient signal to noise ratio and resolution under sub-millisecond sampling. Instead of using traditional point-by-point artificial cold end compensation, an automatic cold end compensation bridge is integrated in the circuit to correct the environmental temperature drift in real time. In addition, in order to suppress crosstalk, diode isolation or micro electro mechanical switch isolation is used at the matrix intersection, ensuring that each thermoelectric junction is disconnected from the readout circuit when not selected, thereby eliminating the signal aliasing problem caused by dense wiring. Optical / wireless readout options: for special environments such as strong electromagnetic interference or ultra-high temperature where metal leads cannot be used, optical or wireless signal transmission schemes are provided. For example, the electrical signal of the micro thermocouple can be converted into an optical signal: using an optical fiber as a transmission medium, the voltage change of each sensing node modulates a low-power laser (intensity or interference modulation), and all optical signals are transmitted out of the high temperature area in different wavelengths / time division multiplexing. For example, in the medium temperature region (<300°C), a micro wireless radio frequency (RF) tag can be integrated, which is powered by the thermocouple and transmits temperature coded signals to realize non-contact reading. The reliability and real-time performance of data acquisition are guaranteed, and even in extreme environments, uninterrupted monitoring can be achieved. Ultra-fast response: through micro / nano scale sensing film and junction, the intrinsic thermal inertia is extremely low, and sub-millisecond or even nanosecond level temperature changes can be captured. The thin film thermocouple sensor can detect a sudden 200°C temperature rise in less than 10 nanoseconds. The thin film / micro wire structure makes the sensor almost "instantaneous" follow the temperature change. In addition, the volume of the sensing points in the array is small and the heat capacity is low, plus direct electrical signal conversion, avoiding thermal hysteresis and complex signal processing delay. Combining these measures, real-time capture of rapid temperature changes can be realized, such as combustion transients, explosion shock or temperature field dynamics in radio frequency ablation surgery of human organs, providing temperature data with sub-millisecond time resolution for these transient processes.

[0017] The temperature field is reconstructed and monitored directly by physical models and electronic circuits: real-time data calibration and linearization: the collected thermal voltages of each point are first passed through the embedded calibration circuit, and the voltage value is converted to the temperature value according to the pre-calibrated material thermoelectric characteristic curve. Due to the use of multiple materials of thermocouple, the corresponding multi-section calibration curve is stored in the memory, and can be automatically switched according to the sensor type. All conversion processes are completed in real time through hardware lookup table or polynomial approximation circuit, without complex calculation. For a wide temperature range, a segmented linearization strategy is adopted to ensure excellent measurement accuracy from low to high temperature segments. Temperature field reconstruction algorithm: after obtaining the discrete point temperature, the system needs to reconstruct it into a continuous temperature field mapping. The traditional scheme uses machine learning for interpolation fitting, while the present application uses a deterministic algorithm and hardware implementation. For example, inverse distance weighted interpolation (IDW) or spline interpolation is applied to quickly calculate the temperature at any position in space from the values of multiple adjacent sensor points. This algorithm can be implemented by FPGA or signal processing ASIC in parallel circuit, and hundreds of point interpolation operations can be completed in microseconds, outputting a grid-ized temperature field matrix signal. In addition, physical models can be introduced for specific applications: for example, in solid material internal temperature measurement, interpolation extrapolation can be based on the analytical solution of the heat conduction equation; in fluid temperature field monitoring, the fluid dynamics model is combined to improve the accuracy of reconstruction. All these calculation processes are hardware logic and firmware programs, and the results are interpretable and have deterministic reliability. Data flow and visualization: the data processing link is highly optimized, from the sensor to the temperature field image using a pipeline structure. The massive data sent by the sensor matrix enters the FPGA through the high-speed ADC, and the calibration, filtering, interpolation and other processes are completed in parallel in the FPGA, generating a temperature field matrix in an instant. The temperature field data can then be transmitted to the host computer or display module, and real-time pseudo-color isotherm, three-dimensional temperature distribution and other intuitive visual output are presented. Since there is no lengthy algorithm iteration, the refresh presentation truly reaches "real-time" (frame rate up to hundreds of Hz). For critical monitoring, threshold monitoring can also be set in hardware: once the temperature in a certain area exceeds the set threshold, the system immediately triggers an alarm or executes control (closed-loop cooling / interlock control: trigger spray cooling, power regulation, mechanical switch) through hardware, and the response time only depends on the electronic circuit delay (microsecond level).

[0018] The multi-point distributed thermocouple array adopts a five-stage composite process of "MEMS microstructure pre-processing - multi-target magnetron sputtering / electron beam evaporation - sol-gel 3D direct writing - laser local cladding - rapid annealing + ALD surface protection", and the specific steps and test standards are as follows:

[0019] MEMS microstructure pre-processing a. Substrate: thickness 25 pm polyimide-alumina composite film; b. Process: deep reactive ion etching (DRIE) to form 100-500 nm deep etch grooves and isolation islands; minimum lithography line width 5 pm, sidewall angle > 87°; c. Quality control: etch depth uniformity < ± 5%, sidewall roughness < 10 nm (AFM), substrate thermal warpage < 50 pm.

[0020] Multi-target magnetron sputtering and e-beam evaporation zoned deposition a. Chamber base pressure , substrate temperature 150-250 °C; b. Low temperature zone deposition Thin film 600 nm, sputtering rate ; c. Medium temperature zone deposition Ni-Si seed layer 200 nm; d. High temperature zone co-deposition W-26%Re and Pt-13%Rh thin films each 3 pm, bias voltage -50 V; e. Quality control: thin film stress < ± 100 MPa (bending test), composition uniformity < ± 2 at% (EDS), roughness Ra < 5 nm.

[0021] Sol-gel 3D direct write wiring a. Nozzle diameter 2.5 pm, printing speed ; b. Printed line width 8 pm ± 0.5 pm, thickness 5-10 pm; c. Post-sintering 400 °C, ; d. Quality control: resistivity , via fill rate > 98% (X-CT).

[0022] Laser localised cladding to build thermoelectric junction a. Laser wavelength 355 nm, pulse width 10 ns, power 25 mW; b. Scan rate , 30 pm x 30 pm area to fuse dual material thin films; c. Quality control: thermoelectric junction resistance < 10 pW, alloy diffusion band width < 5 pm (SEM-EDS).

[0023] Rapid annealing and ALD surface protection a. W-26%Re / Pt-13%Rh zone: 1800 °C, 15 min, argon rapid annealing; b. Bi2Te3 zone: 350 °C, 30 min, nitrogen annealing; c. ALD to cover 100 nm Y2O3 + 200 nm SiO2 composite layer; d. Quality control: thin film stress after annealing < ± 50 MPa, ALD leakage current .

[0024] High-speed row-column strobed acquisition: matrix addressing: each row and column of the array is provided with a 250 pm wide copper-nickel lead, which is cut into 2048 x 2048 temperature measurement pixels; the rows and columns are driven by an N = 2048 3-terminal GaN-HEMT switch array, polled at a 40 MHz clock, and the complete frame reading time is 51.2 ps. Signal conditioning: an LNA (input noise ) + 18-bit 10 MSPS SAR-ADC is arranged on a local amplification board for each 64 x 64 pixel subarray; then it is sent to the FPGA in parallel through an LVDS link. Linearization and compensation: the FPGA performs piecewise fitting (fifth-order polynomial) of the thermoelectric potential-temperature curve of different materials by table lookup method, with an error of ≤±0.1°C. Real-time interpolation: inverse distance weighted (IDW) interpolation is implemented by using a hardware pipeline, with a weight index a = 2; 1 x equivalent high-resolution grid points are generated for every 8 x 8 temperature measurement pixels, with a full-frame update rate of 1 kHz.

[0025] Optical / wireless signal link: optical readout: high-temperature zone voltage signal-analog voltage control micro-ring resonant modulator, 1550 nm optical carrier transmitted through a single-mode optical fiber; after demodulation, the error is <±0.2°C. Wireless readout: medium-temperature zone RF-tag operating frequency 1.6 GHz, thermoelectric voltage modulated carrier phase by a Delta-Sigma ADC, off-power consumption <2 pW.

[0026] Application examples: engine turbine blade: after spraying a heat-resistant oxidized ceramic on the outer wall of the blade, the array is attached, and vacuum ground ignition tests show that the maximum temperature is 1540°C, the hot spot radius is 0.8 mm, and the response time is 0.4 ms. Radio frequency ablation needle: a 32 x 32 embedded array is implanted on the outer wall of a 1.2 mm diameter nickel-titanium tube, and in an in vitro experiment on a pig liver, the diffusion trajectory of the 55°C isothermal surface with power changes is completely characterized within 30 s.

Claims

1. A multi-point distributed thermocouple array, characterized in that: The array includes: Multi-material gradient thermocouple array unit, which forms >1000 points on the substrate The thermoelectric junction matrix is ​​divided into three sections, each using a first thermoelectric material pair covering a temperature range of -200°C to 0°C, a second thermoelectric material pair covering a temperature range of 0°C to 900°C, and a third thermoelectric material pair covering a temperature range of 900°C to 2000°C, to achieve full-temperature range measurement from -200°C to 2000°C. The cold junction constant temperature reference zone has a temperature that is stable at a set reference value and is electrically connected to all cold junctions of the thermocouple array unit through a thermally conductive wire. The high-speed row and column gating signal acquisition module includes a row and column gating switch matrix for thermocouple array units, a low-noise amplifier and a high-resolution analog-to-digital converter, and is configured to complete the potential reading of all nodes of the thermocouple matrix within ≤1ms. The hardware deterministic temperature field reconstruction module includes a lookup table linearization circuit and an inverse distance weighted interpolation circuit, configured to output a rasterized temperature field matrix in real time after receiving and reading data. An optical and / or wireless signal transmission link configured to transmit the read data to a safe area via optical or radio frequency signals in an environment with a temperature greater than 900°C or in the presence of electromagnetic interference; The visualization and alarm interface is configured to output an alarm signal and display the temperature field matrix in real time when the temperature exceeds the limit.

2. The multi-point distributed thermocouple array according to claim 1, characterized in that: The first thermoelectric material pair is Bi2Te3 / Ag, suitable for -200°C to 0°C; the second thermoelectric material pair is Ni-Cr-Si / Cu-Ni, suitable for 0°C to 900°C; the third thermoelectric material pair is tungsten-26% rhenium / platinum-13% rhodium, suitable for 900°C to 2000°C.

3. A multi-point distributed thermocouple array according to claim 1, characterized in that: The thermocouple array unit is fabricated using a MEMS-sputtering-3D direct writing composite process, and the thermocouple unit area is... .

4. A multi-point distributed thermocouple array according to claim 1, characterized in that: The row and column gating switch matrix is ​​a GaN-HEMT board-level array with a driving clock ≥40MHz and a crosstalk rejection ratio ≥80dB.

5. A multi-point distributed thermocouple array according to claim 1, characterized in that: The interpolation circuit of the hardware deterministic temperature field reconstruction module adopts an FPGA parallel pipeline to achieve a frame rate of ≥1kHz and a frame delay of ≤50µs.

6. A multi-point distributed thermocouple array according to claim 1, characterized in that: The optical signal transmission link is completed through 1550nm wavelength division multiplexing optical fiber with a data bandwidth of ≥10Gbps; the wireless signal transmission link adopts 1.6GHz backscatter and the transmission distance is ≤50cm.

7. A multi-point distributed thermocouple array according to claim 1, characterized in that: The visualization and alarm interface outputs a TTL alarm with a delay of <10µs and simultaneously sends a Modbus-TCP control word when the temperature exceeds a set threshold.

8. A real-time temperature field monitoring method for a multi-point distributed thermocouple array as described in any one of claims 1-7, characterized in that, Includes the following steps: S1. Deploy a multi-material gradient thermocouple array unit on the surface or inside of the object to be tested, and connect it to the cold junction constant temperature reference area. S2. Start the high-speed row and column gating signal acquisition module to read the output potential of all thermoelectric junctions at a period of ≤1ms; S3. In the hardware deterministic temperature field reconstruction module, the read potential is linearized and interpolated, and the temperature field matrix is ​​output. S4. Transmit the temperature field matrix to an external visualization and alarm interface via optical and / or wireless transmission links; S5. When the temperature at any point in the temperature field matrix exceeds the preset threshold, an alarm or control signal is output.

9. The real-time temperature field monitoring method according to claim 8, characterized in that: The linearization transformation in step S3 uses a fifth-order polynomial lookup table method, with a single-point transformation time ≤100ns.

10. The real-time temperature field monitoring method according to claim 9, characterized in that: The interpolation process in step S3 uses an inverse distance weighted interpolation algorithm with a weight index of 2 and an interpolation range of no more than 5×5 adjacent temperature measurement pixels.

Images