A capacitive dew point sensor acquisition device and a dew point measurement method

Abstract

This invention relates to the fields of sensor technology and industrial measurement technology, and discloses a capacitive dew point sensor acquisition device. The device includes a sensor capacitor, a conductive connector, and a acquisition circuit board, which are sequentially electrically connected and located within a housing. The conductive connector is located within a sensor base, and polymer sealant is filled between the sensor base and the conductive connector to create a high-pressure, airtight isolation zone. The acquisition circuit board is designed with a stepped structure to be close to the front-end sensor capacitor. A dew point measurement method is also provided. This method uses multi-phase charging and discharging and differential operations to obtain the net discharge time, then establishes a reference dew point using a physical baseline model. Subsequently, a residual regression model is used to compensate for minor nonlinearities, and then post-processing physical hard constraints are performed to output the final dew point. This invention achieves high-pressure leakage and corrosion prevention, suppresses noise interference from parasitic capacitance, and improves the accuracy of dew point measurement over a wide temperature and low humidity range through an inversion algorithm.

Description

Technical Field

[0001] This invention belongs to the fields of sensor technology and industrial measurement technology, specifically relating to a capacitive dew point sensor acquisition device and a dew point measurement method. Background Technology

[0002] Dew point temperature is a key parameter characterizing the humidity state of a gas. It refers to the temperature at which air, under constant water vapor content and pressure, cools to saturation and begins to condense into liquid water. In industrial production and scientific research fields such as compressed air drying, lithium battery manufacturing, natural gas transportation, and meteorological monitoring, dew point temperature is a core indicator for measuring the degree of gas dryness and ensuring process quality and equipment safety.

[0003] Currently, capacitive dew point meters are widely used as the mainstream detection method in industrial settings. However, in practical high-end applications, existing capacitive dew point sensor devices still face the following serious technical challenges:

[0004] Risk of seal failure under high pressure: Industrial compressed air or gas transmission pipelines typically operate under high pressure environments, often reaching 1.6 MPa or even higher. Existing sensor probes often use traditional mechanical interference fits or simple potting processes in their structural packaging. Under long-term high temperature and pressure, these sealing structures struggle to maintain stable airtightness, making it highly susceptible to leakage of the measured high-pressure gas along the probe pins or assembly gaps into the downstream circuit cavity. Gas leakage can not only corrode delicate internal electronic components, leading to measurement failure, but may also cause serious safety accidents.

[0005] Signal transmission interference and parasitic capacitance effects: The capacitance change of capacitive humidity sensors is extremely small, typically in the picofarad range, making them highly sensitive to interference. Traditional sensor designs often suffer from unreasonable structural layouts, such as excessive physical distance between the humidity sensing element and the signal processing circuitry. This physical defect leads to the introduction of significant parasitic capacitance through the leads, severely masking the weak effective capacitance signal and drastically reducing the signal-to-noise ratio and measurement accuracy of the acquisition system.

[0006] Limitations of Dew Point Inversion Algorithms: The conversion from capacitance value to dew point temperature involves complex nonlinear thermodynamic equations. To reduce hardware computing costs, traditional low-cost transmitters typically employ simplified linear approximation algorithms or low-precision lookup table methods. While this approach simplifies calculations, it generates significant cumulative calculation errors in the low-humidity, low-dew-point range where nonlinear characteristics are pronounced, failing to meet the demands of high-precision industrial measurements.

[0007] In summary, there is an urgent need to develop a new integrated data acquisition device that can withstand high temperature and high pressure environments in its mechanical structure, effectively suppress parasitic interference in its hardware layout, and achieve high-precision dew point inversion in its measurement method. Summary of the Invention

[0008] The technical problem to be solved by the present invention is to provide a capacitive dew point sensor acquisition device and a dew point measurement method for dew point measurement under high temperature and high pressure environment, which overcomes the parasitic capacitance and environmental noise interference on the analog signal path, and realizes high-precision calculation of dew point temperature inversion.

[0009] To solve the above-mentioned technical problems, the present invention provides a capacitive dew point sensor acquisition device, comprising: a protective cover, a sensor base and a main body shell connected in sequence from front to back, a sensor capacitor C1 provided inside the protective cover, a conductive connector provided inside the sensor base, and an acquisition circuit board provided inside the main body shell.

[0010] The data acquisition circuit board includes sensor connection pads, a humidity detection module, a temperature detection module, and a microcontroller. The two ends of the conductive connector are electrically connected to the sensor connection pads and the sensor capacitor C1, respectively. The humidity detection module includes a sensing measurement channel and a reference measurement channel. The sensing measurement channel is connected to the sensor connection pads and the microcontroller signal; the reference measurement channel is connected to the microcontroller signal; and the temperature detection module is connected to the microcontroller signal.

[0011] The sensor connection pad is located at the very front of the acquisition circuit board. The humidity detection module and temperature detection module are located close to the sensor connection pad. The gap between the sensor base and the conductive connector is filled with polymer sealant.

[0012] As an improvement to the capacitive dew point sensor acquisition device of the present invention:

[0013] The sensor base is connected to the protective cover and the main body shell at both ends by threads, and a sealing element is provided between the protective cover and the sensor base, and between the sensor base and the main body shell.

[0014] As a further improvement to the capacitive dew point sensor acquisition device of the present invention:

[0015] The sensing and measurement channel includes a first analog switch matrix U1. The VCC power supply is connected to the power input terminal of the first analog switch matrix U1 through a charging resistor R1. The first electrode and the second electrode of the sensor capacitor C1 are respectively connected to the channel selection terminal of the first analog switch matrix U1.

[0016] The reference measurement channel includes a second analog switch matrix U2 and an onboard reference capacitor. The VCC power supply is connected to the power input terminal of the second analog switch matrix U2 via charging resistor R3; the reference capacitor... The first and second electrodes are respectively connected to the channel selection terminals of the second analog switch matrix U2;

[0017] The microcontroller includes an MCU chip U3, which is connected to the first analog switch matrix U1 and the second analog switch matrix U2 via an SPI communication bus. The SPI communication bus includes a common clock signal line (CLK), a master-in-slave-out signal line (MISO), and a master-out-slave-in signal line (MOSI).

[0018] The first chip select terminal of MCU chip U3 is connected to the synchronization signal terminal of the first analog switch matrix U1, and the first signal acquisition terminal is connected to the sensing measurement terminal of the first analog switch matrix U1 as a sensing measurement bus (MCU_COM).

[0019] The second chip select terminal of MCU chip U3 is connected to the synchronization signal terminal of the second analog switch matrix U2, and the second signal acquisition terminal is connected to the reference measurement terminal of the second analog switch matrix U2 as a reference measurement bus (MCU_COM1).

[0020] The temperature detection module includes a voltage divider network, which is composed of an NTC thermistor and a precision reference resistor connected in series. The VCC power supply is grounded after passing through the voltage divider network. The common terminal of the NTC thermistor and the precision reference resistor is connected to the analog-to-digital conversion input channel of the MCU chip U3.

[0021] The present invention also provides a method for measuring dew point using a capacitive dew point sensor acquisition device, comprising the following specific steps:

[0022] Step 1: The microcontroller constructs analog capacitor detection circuits using the first analog switch matrix U1 and the second analog switch matrix U2, respectively;

[0023] Then, the sensor capacitor C1 is continuously sampled. During each sampling, the sensing measurement channel is selected first, and multi-phase charge-discharge measurement and differential operation are performed on the sensor capacitor C1 to obtain the net discharge time T. net Then select the reference measurement channel for the onboard reference capacitor. Perform multi-phase charge-discharge measurements and differential calculations to obtain the reference time. Then, the absolute capacitance value of the sensor is obtained by using the ratio measurement principle. ;

[0024] Step 2: Set the absolute capacitance value of the sensor Filtering is performed to obtain the smoothed capacitance value. ;

[0025] Step 3: Synchronously collect the ambient temperature T, and calculate the reference dew point temperature using a preset physical baseline model. ;

[0026] Step 4: Based on smoothing capacitance value Ambient temperature T and reference dew point temperature Constructing Physically Extended Eigenvectors Physically extended feature vectors The input is fed into a pre-trained residual regression model, and the output is the predicted residual value. ;

[0027] Step 5: Predict the residual values Perform residual amplitude truncation to obtain truncated residuals ; set the reference dew point temperature With truncated residuals Combined, the final dew point temperature is calculated. .

[0028] As an improvement to the dew point measurement method of the present invention:

[0029] The multi-phase charge-discharge measurement process includes:

[0030] First phase measurement: Construct a measurement circuit with sensor capacitance C1 and parasitic capacitance Cp1 in parallel and charge and discharge it, and record the first discharge time T1;

[0031] Second phase measurement: Construct a measurement circuit with sensor capacitance C1 and parasitic capacitance Cp2 in parallel and charge and discharge it, and record the second discharge time T2;

[0032] Third phase measurement: The two ends of the sensor capacitor C1 are shielded at the same potential, and a measurement circuit containing only the parasitic capacitances Cp1 and Cp2 in parallel is constructed and charged and discharged. The third discharge time T3 is recorded.

[0033] The formula for calculating the difference in net discharge time is: .

[0034] As a further improvement to the dew point measurement method of the present invention:

[0035] The absolute capacitance value of the sensor The calculation formula is:

[0036] in, For the onboard reference capacitor The reference time is obtained by performing the multi-phase charge-discharge measurement and the differential calculation.

[0037] The smoothing capacitance value The calculation formula is: ,in, The absolute capacitance value of the sensor is obtained in the nth acquisition, and N is the preset capacity of the static capacitor array.

[0038] As a further improvement to the dew point measurement method of the present invention:

[0039] The physical baseline model is as follows:

[0040]

[0041] Where C0 is the zero-point capacitance background value of the sensor under absolutely dry conditions; parameters a, b, and c are system calibration coefficients determined in advance by least squares fitting.

[0042] The physical extended feature vector for:

[0043] .

[0044] As a further improvement to the dew point measurement method of the present invention:

[0045] The formula for the residual amplitude truncation process is expressed as follows:

[0046]

[0047] Where M is the maximum nonlinear deviation threshold;

[0048] The formula for calculating the final dew point temperature is as follows:

[0049] .

[0050] As a further improvement to the dew point measurement method of the present invention:

[0051] The analog capacitor detection circuit constructed through the first analog switch matrix U1 includes gating switches A0 and A1, grounding switches B0 and B1, charging switch D1, and discharging switch D0.

[0052] The VCC power supply terminal is connected to the sensor measurement bus via a charging resistor and a charging switch D1. The sensor measurement bus is connected to ground (GND) via a discharge switch D0 and a discharge resistor.

[0053] The first electrode of sensor capacitor C1 is connected to one end of selector switch A1 and one end of ground switch B1, respectively. The other end of selector switch A1 is connected to the sensor measurement bus, and the other end of ground switch B1 is connected to ground (GND). The second electrode of sensor capacitor C1 is connected to one end of selector switch A0 and one end of ground switch B0, respectively. The other end of selector switch A0 is connected to the sensor measurement bus, and the other end of ground switch B0 is connected to ground (GND).

[0054] In the measurement state, there is an equivalent parasitic capacitance Cp1 between the first electrode of sensor capacitor C1 and ground, and an equivalent parasitic capacitance Cp2 between the second electrode of sensor capacitor C1 and ground (GND). The microcontroller controls the switching on and off of gate switches A0 and A1, ground switches B0 and B1, and charging and discharging switches D1 and D0 to connect sensor capacitor C1, parasitic capacitance Cp1, and parasitic capacitance Cp2 to the sensing measurement bus and perform charging and discharging operations.

[0055] The beneficial effects of this invention are mainly reflected in:

[0056] 1. This invention constructs an absolute isolation zone capable of withstanding high pressure for extended periods by filling and curing polymer sealant into the gaps of the conductive connectors in the sensor base. This completely eliminates leakage of the measured high-pressure gas into the rear circuit cavity along the pin gaps. Furthermore, the stepped, integrated design of the narrow front section of the acquisition circuit board allows the humidity detection module to be placed as close as possible to the front-end humidity-sensitive element. This "zero-lead" PCB layout significantly shortens the analog signal transmission path from a physical perspective and achieves thermal isolation between the heat source and the sensitive source, improving the thermal stability of the measurement.

[0057] 2. This invention utilizes a microcontroller and an analog switch matrix in synergy to construct different equivalent detection loops in a time-series manner without adding additional complex impedance measurement hardware. Then, by employing the ratio measurement principle and differential compensation calculations of multi-phase charge / discharge times (T1, T2, T3), the system can dynamically cancel the inherent parasitic capacitances (Cp1, Cp2) of pins and PCB traces to ground at the underlying physical link, and eliminate common-mode errors caused by power supply fluctuations and resistor temperature drift. This mechanism ensures that the final extracted net discharge time accurately and uniquely characterizes the change in the sensor's own capacitance, significantly improving the robustness and signal-to-noise ratio of small capacitance data acquisition.

[0058] 3. To address the technical shortcomings of existing low-cost transmitter lookup table methods, such as low accuracy, and the tendency of pure data-driven models (such as standard SVR) to overfit, the dew point detection method of this invention reduces the computational complexity of high-dimensional mapping and achieves high-precision and high-stability dew point inversion. Attached Figure Description

[0059] The specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0060] Figure 1 This is a schematic diagram of the capacitive dew point sensor acquisition device of the present invention;

[0061] Figure 2 This is a schematic diagram of the layout of the acquisition circuit board of the present invention;

[0062] Figure 3This is a schematic diagram illustrating the functional principle of the data acquisition circuit board of the present invention;

[0063] Figure 4 This is a schematic diagram of the dew point detection and measurement process of the present invention;

[0064] Figure 5 This is an analog capacitance detection circuit for the charging phase during the first phase measurement of the sensing measurement channel of the present invention.

[0065] Figure 6 This invention provides an analog capacitance detection circuit for the discharge and timing stages during the first phase measurement of the sensing measurement channel of the present invention.

[0066] Figure 7 This is an analog capacitance detection circuit for the charging phase during the second phase measurement of the sensing measurement channel of the present invention.

[0067] Figure 8 This invention provides an analog capacitance detection circuit for the discharge and timing stages during the second phase measurement of the sensing measurement channel.

[0068] Figure 9 This is an analog capacitance detection circuit for the charging phase during the third phase measurement of the sensing measurement channel of the present invention.

[0069] Figure 10 This invention provides an analog capacitance detection circuit for the discharge and timing stages during the third phase measurement of the sensing measurement channel in the present invention.

[0070] Figure 11 This is a cross-sectional schematic diagram of the sensor base and conductive connector of the present invention;

[0071] Figure 12 This is a schematic diagram of the circuit topology of the sensing and measurement channel of the present invention;

[0072] Figure 13 This is a schematic diagram of the circuit topology of the reference measurement channel of the present invention;

[0073] Figure 14 This is a schematic diagram of the circuit topology of the MCU chip of the present invention. Detailed Implementation

[0074] The present invention will be further described below with reference to specific embodiments, but the scope of protection of the present invention is not limited thereto:

[0075] Example 1: A capacitive dew point sensor acquisition device, such as... Figure 1 As shown, the structure is cylindrical and includes, from front to back along the axial direction, a protective cover 1, a sensor base 2, a main body shell 7, a tail fixing sleeve 8, and an electrical interface assembly 9.

[0076] The protective cover 1 is a cylindrical structure with one end closed and the other end threaded. It is preferably made by stainless steel powder metallurgy sintering process to form a porous structure with micron-sized pores. This is intended to effectively block external particulate impurities from entering, while ensuring that the gas being measured can quickly penetrate and contact the front-end humidity-sensitive element inside.

[0077] The sensor base 2 has external threads on both the front and rear sides, and its inner cavity is hollow, such as... Figure 11 As shown, the sensor base 2 is connected to the protective cover 1 and the main body shell 7 at both ends by threads. At the same time, sealing rings are provided between the protective cover 1 and the sensor base 2, and between the sensor base 2 and the main body shell 7. The sealing rings are fitted on the sensor base 2 to achieve airtight sealing.

[0078] The tail fixing sleeve 8 is inserted into the main body housing 7. A set screw is screwed into the radial threaded hole at the end of the main body housing 7 to lock and fix it with the tail fixing sleeve 8, thereby preventing it from axially dislodging and circumferentially rotating. Both the tail fixing sleeve 8 and the electrical interface assembly 9 are hollow and are used to collect the wiring of the circuit board 6 to external electrical connections.

[0079] A sensor capacitor C1 is housed inside the protective cover 1 as a front-end humidity-sensitive element. A conductive connector 4 is located inside the sensor base 2, and a data acquisition circuit board 6 is located inside the main housing 7. The conductive connector 4 consists of two metal pins. The main body of the metal pins passes through the inner cavity of the sensor base 2, with the front end extending into the protective cover 1 and fixedly connected to the two ends of the sensor capacitor C1 by welding. The rear end extends into the main housing 7 and is fixedly connected to the data acquisition circuit board 6 by welding. This allows the sensor capacitor C1 and the data acquisition circuit board 6 to establish a signal connection, enabling it to be used as a humidity sensor.

[0080] To address the technical challenges of leakage and insulation failure under high-pressure conditions in existing technologies, the gap between the sensor base 2 and the conductive connector 4 is filled with a high-temperature and high-pressure resistant polymer sealant. After curing, this sealant isolates the inner cavity of the protective cover 1 and the inner cavity of the main body shell 7 into two sealed areas. This allows the inner cavity of the main body shell 7 to withstand gas pressure for extended periods, completely preventing the high-pressure gas being measured from leaking along the gaps in the conductive connector 4 into the rear inner cavity of the main body shell 7. The polymer sealant exhibits excellent high-temperature resistance, particularly withstanding the high-temperature impact generated during welding during assembly. When the rear end of the conductive connector 4 is welded to the acquisition circuit board 6 at high temperatures, even if heat is conducted along the conductive connector 4 to the sealed area, the polymer sealant maintains a stable physical state, without softening, expanding, or carbonizing. This prevents sealing failure or pin displacement caused by welding thermal stress, while simultaneously ensuring long-term electrical insulation between the metal pin and the sensor base 2.

[0081] A pair of guide grooves are symmetrically formed along the axial direction on the inner wall of the main body shell 7 for inserting and fixing the acquisition circuit board 6. The acquisition circuit board 6 adopts an integrated design adapted to the internal space of the main body shell 7, and has a stepped structure that gradually narrows from back to front. During assembly, the narrow end of the acquisition circuit board 6 faces forward and is inserted into the main body shell 7 from back to front, with the narrow end of the acquisition circuit board 6 abutting against the front end of the main body shell 7. Then, the tail fixing sleeve 8 is inserted into the tail of the main body shell 7, with the front end of the tail fixing sleeve 8 abutting against the tail end of the acquisition circuit board 6, thereby restricting the axial movement of the acquisition circuit board 6. The tail fixing sleeve 8 and the electrical interface assembly 9 are threaded together. The tail end of the electrical interface assembly 9 is provided with a rubber elastic sealing ring to achieve a seal between the external cable and the electrical interface assembly 9, preventing dust, moisture or rainwater from the external environment of the industrial site from flowing back into the circuit cavity.

[0082] The acquisition circuit board 6 is divided along the axial direction into a wide tail section, a middle control section, and a narrow front section, as shown below. Figure 2 As shown.

[0083] The rear wide-body area is used for interface and power management, and houses the electrical interface 100. The electrical interface 100 includes and integrates an external +24V power input pin, a 4-20mA analog output pin, and an RS-485 digital communication pin. Adjacent to the electrical interface 100 is a power management module, including a DC-DC step-down module 101 and an LDO linear regulator 102.

[0084] The central control area is used for main control and output drive, and is equipped with a 4-20mA module 103, an RS-485 module 104, and a microcontroller 105 as the core control unit.

[0085] The core humidity detection module 106 and temperature detection module 107 are arranged in the narrow front section, and are positioned close to the sensor connection pad 108 at the very front of the circuit board. The sensor connection pad 108 is used to solder to the tail end of the conductive connector 4, minimizing the physical distance between the humidity detection module 106 and the front-end humidity sensing element (sensor capacitor C1). This "zero-lead" layout design significantly reduces parasitic capacitance and environmental noise interference on the analog signal path from a physical perspective, providing the hardware foundation for achieving high-precision measurement.

[0086] The circuit function principle of acquisition circuit board 6 is as follows: Figure 3 As shown, a complete electrical circuit is constructed using the microcontroller 105 as the core, comprising three main electrical units: power management, signal acquisition, and output control. The specific connection relationships and working principles of each functional unit are as follows:

[0087] Power Management Unit: The +24V power input pin of the electrical interface 100 is electrically connected to the input terminal of the DC-DC step-down module 101; the output terminal of the DC-DC step-down module 101 is electrically connected to the input terminal of the LDO linear regulator 102; the output terminal of the LDO linear regulator 102 serves as the VCC power supply and is electrically connected to the power input terminals of the microcontroller 105 and the humidity detection module 106, respectively. In operation, the external +24V power supply, after being input through the electrical interface 100, first connects to the DC-DC step-down module 101, efficiently reducing the input voltage to 5V; then it connects to the LDO linear regulator 102, further converting the 5V voltage into a 3.3V system voltage with extremely low ripple, providing a stable VCC power supply for the microcontroller 105 and the humidity detection module 106.

[0088] Signal acquisition unit: includes humidity detection module 106, temperature detection module 107, and reference capacitor. .

[0089] The humidity detection module 106 adopts a dual-channel differential symmetrical architecture, including a physically independent sensing measurement channel and a reference measurement channel. The sensing measurement channel is connected to the front-end humidity-sensitive element (i.e., sensor capacitor C1) via sensor connection pad 108 and conductive connector 4. The reference measurement channel is connected to the reference capacitor. Signal connection, reference measurement channel detects reference capacitance This serves as the detection benchmark for the sensor capacitance C1.

[0090] In this embodiment, the microcontroller 105 includes, preferably, an MCU chip U3 of model N32G430F8S7, such as... Figure 14 As shown. The sensing measurement channel includes a first analog switch matrix U1 (preferably ADG1414BCPZ), a charging resistor R1 (30kΩ), and a discharging resistor R2 (30kΩ), as follows. Figure 12 As shown. The VCC power supply output by the LDO linear regulator 102 is connected to the power input terminal (pin S1) of the first analog switch matrix U1 through the charging resistor R1. The first and second electrodes of the sensor capacitor C1 are respectively connected to the corresponding channel selection terminals of the first analog switch matrix U1: the first electrode of the sensor capacitor C1 is connected to pins D2 and S3 of the first analog switch matrix chip U1, and the second electrode is connected to pins D6 and S5 of the first analog switch matrix U1.

[0091] The first chip select pin (PA4) of MCU chip U3 is connected to the synchronization signal pin (SYNC) of the first analog switch matrix U1 for communication of the first chip select signal CS. Pin PA5 of MCU chip U3 is connected to the SCLK pin of the first analog switch matrix U1 as the clock signal line (CLK). Pin PA6 of MCU chip U3 is connected to the SDO pin of the first analog switch matrix U1 as the master-in, slave-out signal line (MISO). Pin PA7 of MCU chip U3 is connected to the DIN pin of the first analog switch matrix U1 as the master-out, slave-in signal line (MOSI). The first signal acquisition pin (PA0) of MCU chip U3 is connected to the sensing measurement pin (short-circuited pins D1, S2, S4, and D5) of the first analog switch matrix U1 as the sensing measurement bus (MCU_COM). The first analog switch matrix U1, under the control of the microcontroller 105's SPI communication bus (including clock signal line (CLK), master input slave output signal line (MISO), master output slave input signal line (MOSI)) and the first chip select signal CS, constructs a charging and discharging circuit for sensor capacitor C1, and feeds back the signal to the microcontroller 105's signal acquisition channel (PA0 pin) in real time through the sensor measurement bus (MCU_COM).

[0092] The circuit topology, components, and component models of the reference measurement channel are consistent with those of the sensing measurement channel, such as... Figure 13 As shown, it includes a second analog switch matrix U2 (preferably ADG1414BCPZ) and an onboard reference capacitor. Charging resistor R3 (30kΩ) and discharging resistor R4 (30kΩ). The onboard reference capacitor... Made of C0G ceramic dielectric, with a nominal capacitance of 30pF. The electrical connection of the reference measurement channel is as follows:

[0093] The VCC power supply output from the LDO linear regulator 102 is connected to the power input terminal (pin S1) of the second analog switch matrix U2 via charging resistor R3. Reference capacitor. The first and second electrodes are respectively connected to the channel selection terminals corresponding to the second analog switch matrix U2; the reference capacitor The first electrode is connected to pins S2 and S3 of the second analog switch matrix chip U2, and the second electrode is connected to pins D1 and D2 of the second analog switch matrix U2.

[0094] The second chip select pin (PA3) of MCU chip U3 is connected to the synchronization signal pin (SYNC) of the second analog switch matrix U2 for communication via the second chip select signal CS1. Pin PA5 of MCU chip U3 is connected to pin SCLK of the second analog switch matrix U2 as the clock signal line (CLK). Pin PA6 of MCU chip U3 is connected to pin SDO of the second analog switch matrix U2 as the master-in, slave-out signal line (MISO). Pin PA7 of MCU chip U3 is connected to pin DIN of the second analog switch matrix U2 as the master-out, slave-in signal line (MOSI). That is, the second analog switch matrix U2 and the first analog switch matrix U1 share the SPI communication bus of MCU chip U3 (MOSI connected to pin PA7, MISO to pin PA6, and SCLK to pin PA5), but receive independent timing drive from microcontroller 105 through a separate second chip select signal CS1 (corresponding to pin PA3).

[0095] Pin PA2 of MCU chip U3 is connected to pins D1 and D2 of the second analog switch matrix U2 as a reference measurement bus (MCU_COM1).

[0096] In addition, such as Figure 14 As shown, the MCU chip U3 of this device is also surrounded by several capacitors (such as C2, C4, C5, C7, C11, etc.), resistors (such as R5), crystal oscillator (U4), and other electronic components. These components constitute the microcontroller 105, the analog switch matrix, and the power supply filtering circuit, clock oscillation circuit, reset configuration circuit, and impedance matching circuit of related integrated chips, etc., which are all conventional circuit designs of the MCU chip U3 (N32G430F8S7). The corresponding component selection and design specifications can be easily obtained from publicly available information and are not within the scope of this invention. Therefore, they will not be described in detail here.

[0097] In this embodiment, the first analog switch matrix U1 and the second analog switch matrix U2 of the sensing measurement channel and the reference measurement channel employ CMOS analog switch chips with low impedance and low charge injection effect (e.g., conventional ADG1414 or multiplexers / analog switches with similar electrical characteristics). Under the strict timing control of the microcontroller 105, and in conjunction with the comparator or timer unit of the microcontroller 105, they jointly realize the time-to-discharge (TDC) conversion of the small capacitor. This customized hardware layout based on discrete switch arrays is the core hardware foundation for this invention to implement three-phase differential measurement and cut off parasitic capacitance interference from the underlying physical link.

[0098] In this embodiment, the microcontroller 105 controls the switch networks inside the first analog switch matrix U1 and the second analog switch matrix U2, effectively constructing an analog capacitance detection circuit including a gating switch A0, a gating switch A1, a grounding switch B0, a grounding switch B1, a charging switch D1, a discharging switch D0, parasitic capacitance Cp1, and parasitic capacitance Cp2. The parasitic capacitances Cp1 and Cp2 are not actual physical discrete devices, but rather equivalent logic parameters characterizing the capacitance distributed to ground of the sensor capacitance C1 pin and the PCB traces. In the measurement state, there is an equivalent parasitic capacitance Cp1 between the first electrode of sensor capacitor C1 and ground, and an equivalent parasitic capacitance Cp2 between the second electrode of sensor capacitor C1 and ground. The microcontroller 105 controls the switching on and off of selector switches A0 and A1, ground switches B0 and B1, charging switch D1, and discharging switch D0 to connect sensor capacitor C1, parasitic capacitance Cp1, and parasitic capacitance Cp2 to the sensor measurement bus (MCU_COM) and perform charging and discharging operations.

[0099] The power supply terminal (VCC) is connected to the sensor measurement bus (MCU_COM) via charging resistor R1 and charging switch D1. Simultaneously, the sensor measurement bus (MCU_COM) is also connected to ground (GND) via discharge switch D0 and discharge resistor R2.

[0100] The first electrode of the sensor capacitor C1 (e.g.) Figure 5 The upper end of the sensor capacitor C1 shown is connected to one end of the gating switch A1, one end of the grounding switch B1, and one end of the parasitic capacitor Cp1, respectively. The other end of the gating switch A1 is connected to the sensor measurement bus (MCU_COM); the other end of the grounding switch B1 is connected to ground (GND); and the other end of the parasitic capacitor Cp1 is connected to ground (GND).

[0101] The second electrode of the sensor capacitor C1 (such as...) Figure 5 The lower end of the sensor capacitor C1 shown is connected to one end of the gating switch A0, one end of the grounding switch B0, and one end of the parasitic capacitor Cp2. The other end of the gating switch A0 is connected to the sensor measurement bus (MCU_COM); the other end of the grounding switch B0 is connected to ground (GND); and the other end of the parasitic capacitor Cp2 is connected to ground (GND).

[0102] Through the above connection relationship, the sensor measurement bus (MCU_COM) is configured to control the on and off states of the charging switch D1, discharging switch D0, gating switch A0, gating switch A1, grounding switch B0 and grounding switch B1 to realize the charging, discharging and sampling operations of the sensor capacitor C1.

[0103] Since the sensing measurement channel and the reference measurement channel are the same, the specific connection implementation of the analog capacitance detection circuit of the second analog switch matrix U2 can be obtained by analogy, that is, in Figures 5-10 In the analog capacitance detection circuit, an onboard reference capacitor is used. Replace sensor capacitor C1, replace charging resistor R1 with charging resistor R3, replace discharging resistor R2 with discharging resistor R4, and replace the sensor measurement bus (MCU_COM) with the reference measurement bus (MCU_COM1). The specific device connection relationships will not be described again here.

[0104] The temperature detection module 107 employs an analog voltage divider sampling topology. The VCC power supply terminal is grounded (GND) after passing through an NTC thermistor (e.g., but not limited to a precision NTC component with a nominal resistance of 10kΩ or 100kΩ and a B-value accuracy error of less than ±1%) and a low-temperature drift precision reference resistor (10kΩ or 100kΩ). The common terminal of the NTC thermistor and the precision reference resistor is connected to the ADC analog input channel of the microcontroller 105. This invention uses a high-precision NTC thermistor and a precision reference resistor connected in series to form a voltage divider network. The analog voltage signal output by this voltage divider network is directly connected to the ADC analog input channel of the MCU chip U3. The MCU chip U3 performs high-speed sampling and nonlinear calibration on this voltage signal to obtain the current ambient dry-bulb temperature value in real time, which is used as one of the input parameters for subsequent dew point measurement methods.

[0105] Output control unit: includes a 4-20mA module 103 and an RS-485 module 104. The 4-20mA module 103 contains a voltage-to-current conversion circuit. The MCU chip U3 outputs an adjustable duty cycle control signal through the PWM interface, which is converted into a standard 4-20mA analog current signal by this module and output through the electrical interface. The RS-485 module 104 contains a transceiver chip. The MCU chip U3 communicates with it through the UART interface to convert the processed dew point data into a differential signal and send it externally through the electrical interface.

[0106] Based on a capacitive dew point sensor acquisition device, this embodiment proposes an innovative dew point inversion method. This method utilizes a microcontroller to execute a series of precise controls and complex numerical calculations, aiming to extract high-precision dew point temperature from the raw signal, which is subject to environmental interference. The entire measurement process is as follows: Figure 4 As shown, the specific steps include:

[0107] Step 1: Capacitance signal acquisition and parasitic elimination

[0108] To eliminate ground-to-ground parasitic capacitance interference introduced by leads, PCB traces, and connectors, the MCU chip U3 controls the capacitance detection circuit to execute floating measurement timing. Multi-phase charge-discharge measurements are performed on the sensor capacitor C1, and differential operations are used to eliminate circuit parasitic capacitance to obtain the original capacitance value. Specifically, the MCU chip U3 drives the analog capacitance detection circuits of the first analog switch matrix U1 and the second analog switch matrix U2 to construct different circuit topologies. Then, multi-phase charge-discharge measurements and differential operations are performed on the sensor capacitor C1 through the sensing measurement channel to obtain the net discharge time T. net Then, the onboard reference capacitor is measured via the reference measurement channel. Perform multi-phase charge-discharge measurements and differential calculations to obtain the reference time. Then, the absolute capacitance value of the sensor is obtained by using the ratio measurement principle. Specifically:

[0109] Step 1.1: First phase measurement: capture the sum of the capacitance of the sensor capacitor C1 and the parasitic capacitance Cp1, as well as the first discharge time T1.

[0110] Step 1.1.1: Initial Circuit Topology Construction:

[0111] MCU chip U3 selects the sensing measurement channel via the first chip select signal CS. MCU chip U3 first controls the analog switch matrix to construct a specific measurement loop for the "parallel combination of sensor capacitance and single-sided parasitic capacitance," such as... Figure 5 As shown. The specific operation is as follows: MCU chip U3 controls the closing of grounding switch B1 via the SPI interface, forcibly pulling the first electrode of sensor capacitor C1 low to ground potential (GND); simultaneously, it closes select switch A0, selecting the lower end of sensor capacitor C1 to connect to the measurement bus (MCU_COM). In this topology, the second electrode of sensor capacitor C1 is connected to the measurement bus (MCU_COM), and the other end is grounded; the inherent parasitic capacitance to ground Cp2 in the circuit is also located between the measurement bus (MCU_COM) and ground. Therefore, in electrical connection, sensor capacitor C1 and parasitic capacitance Cp2 actually form a parallel structure, and the total capacitance to be measured is... The parasitic capacitance Cp1 is short-circuited and shielded, and does not participate in the measurement.

[0112] Step 1.1.2: Charging Stage (refer to...) Figure 5 )

[0113] After the initial topology is built, the MCU chip U3 initiates the charging process. First, the charging switch D1 is opened, while the discharging switch D0 is closed, ensuring that the initial states of sensor capacitor C1 and parasitic capacitance Cp2 are cleared. Then, the discharging switch D0 is opened and the charging switch D1 is closed, injecting current from the power supply voltage (VCC) into the measurement bus (MCU_COM) via the charging resistor R1 and the closed charging switch D1. The current flows into the parallel-connected sensor capacitor C1 and parasitic capacitance Cp2. As charge accumulates, the voltage at the measurement node increases exponentially until it reaches the power supply voltage (VCC) or the preset full-charge voltage, completing the charging operation. The thick black arrows in the diagram indicate the current path.

[0114] Step 1.1.3: Discharge and Timing Stage (Refer to...) Figure 6 )

[0115] After charging is complete, the MCU chip U3 immediately switches to discharge measurement mode. Charging switch D1 is disconnected, and discharge switch D0 is simultaneously closed. The charge stored in sensor capacitor C1 and parasitic capacitance Cp2 begins to discharge to ground through the closed discharge switch D0 and the high-precision discharge resistor R2. At the instant discharge switch D0 is closed, the timer unit inside the MCU chip U3 synchronously starts the comparator trigger signal timing. As the charge discharges, the voltage at the measurement node gradually decreases. When the voltage drops to the preset comparator flip-flop threshold voltage (V... th When the comparator trigger signal stops the timing, the recorded time interval is the first discharge time T1. According to the RC discharge principle (T=R⋅C⋅ln(V)...),... init / V th ), with discharge resistor R2, initial voltage and comparator switching threshold voltage V th Under the condition that all are fixed, the first discharge time T1 is proportional to the time constant RC. Therefore, the first discharge time T1 is linearly proportional to the total parallel capacitance (C1 + Cp2).

[0116] Step 1.2: Second phase measurement: capture the sum of the sensor capacitance and the parasitic capacitance on the other side.

[0117] Step 1.2.1: Circuit Topology Construction

[0118] MCU chip U3 switches the analog switch state to reconstruct the measurement circuit. Specifically, it closes the grounding switch B0, forcibly pulling the second electrode of sensor capacitor C1 low to ground potential (GND); simultaneously, it closes the selection switch A1, selecting the first electrode of sensor capacitor C1 to connect to the measurement bus (MCU_COM). In this topology, sensor capacitor C1 and parasitic capacitance Cp1 in the circuit form a parallel structure connected to the measurement circuit, while parasitic capacitance Cp2 is short-circuited and shielded, not participating in the measurement.

[0119] Step 1.2.2 Charging Stage (refer to...) Figure 7 )

[0120] Similar to the first phase measurement, MCU chip U3 closes charging switch D1 and opens discharging switch D0. Power supply current is injected into the measurement bus (MCU_COM) through charging resistor R1, charging the parallel sensor capacitor C1 and parasitic capacitance Cp1 respectively until the full charge voltage is reached.

[0121] Step 1.2.3 Discharge and Timing Stage (refer to...) Figure 8 )

[0122] After charging is complete, MCU chip U3 disconnects charging switch D1 and closes discharging switch D0. The charge stored in sensor capacitor C1 and parasitic capacitance Cp1 begins to discharge. The high-precision timer inside MCU chip U3 synchronously starts the comparator trigger signal to time and record the duration of the discharge process. This time is the second discharge time T2, and its value is proportional to the total parallel capacitance C1 + Cp1.

[0123] Step 1.3: Third Phase Measurement: Sum of Captured Parasitic Capacitances

[0124] Step 1.3.1: Circuit Topology Construction (refer to...) Figure 9 )

[0125] The MCU chip U3 controls the analog switch to enter a specific "equipotential shielding" state. Specifically, it simultaneously closes selector switches A0 and A1, connecting the upper and lower ends of sensor capacitor C1 to the measurement bus (MCU_COM). In this special topology, the potentials across sensor capacitor C1 remain constant (i.e., in a short-circuit state), thus preventing charge storage; only the inherent parasitic capacitances Cp1 and Cp2 remain in parallel within the measurement circuit.

[0126] Step 1.3.2: Charging Stage (refer to...) Figure 9 ):

[0127] The MCU chip U3 closes the charging switch D1 and opens the discharging switch D0. Current flows only into the parasitic capacitors Cp1 and Cp2, while the sensor capacitor C1 does not participate in the charging process because its two ends are at the same potential.

[0128] Step 1.3.3: Discharge and Timing Stage (Refer to...) Figure 10 ):

[0129] The MCU chip U3 disconnects the charging switch D1 and closes the discharging switch D0. The charge stored in parasitic capacitors Cp1 and Cp2 begins to discharge. The high-precision timer inside the microcontroller 105 synchronously starts the comparator trigger signal to time and records the discharge time at this moment, which is the third discharge time T3, and its value is proportional to the total parallel capacitance Cp1 + Cp2.

[0130] Step 1.4: Difference Calculation (Parasitic Elimination):

[0131] MCU chip U3 performs differential compensation calculations based on the measurement results from steps 1.1, 1.2, and 1.3 above. According to the principle of parallel capacitors, the sum of the first two measurements (T1 + T2) includes twice the sensor signal and all parasitic signals, while the third measurement (T3) includes only the parasitic signals. MCU chip U3 uses the formula... Calculate the net discharge time T net Through this step, the influence of parasitic capacitances (Cp1, Cp2) drifting with the environment can be largely eliminated, and the final extracted net discharge time T net This is a precise representation of the capacitance of the humidity sensor itself.

[0132] Step 1.5: Ratio Calculation and Capacitance Conversion

[0133] To eliminate common-mode errors caused by power supply voltage fluctuations and discharge resistor temperature drift, MCU chip U3 selects the reference measurement channel via the second chip select signal CS1 to measure the onboard reference capacitor. Perform multi-phase charge-discharge measurements and differential calculations according to steps 1.1-1.4 to obtain the reference time. The final absolute capacitance value of the sensor is calculated using the ratio measurement principle. :

[0134]

[0135] sensor absolute capacitance value This refers to the original capacitance data after eliminating parasitic interference and environmental drift.

[0136] Step 2: The MCU chip U3 controls the humidity detection module 106 to continuously sample the front-end humidity-sensitive element (i.e., sensor capacitor C1). To eliminate random noise interference, the system maintains a static capacitor array with a capacity of N (N=80 in this embodiment) in memory and performs recursive moving average filtering on the continuously collected raw capacitance values. The filtered smooth capacitance value... The calculation formula is:

[0137]

[0138] in, This represents the absolute capacitance value of the sensor acquired in the nth acquisition. This step allows for the acquisition of highly stable capacitance measurement data.

[0139] Step 3: The MCU chip U3 uses the smoothing capacitor value obtained in Step 2. The reference dew point temperature is calculated using a semi-empirical physical baseline model (based on the Magnus formula) pre-stored in non-volatile memory (such as Flash) and the synchronously acquired ambient temperature T. The specific calculation formula is as follows:

[0140]

[0141] Where C0 is the zero-point capacitance background value of the sensor under absolutely dry conditions; parameters a, b, and c are system calibration coefficients determined in advance through least squares fitting. This step aims to transform the complex water vapor adsorption mechanics into a directly calculable logarithmic analytical expression, capturing the main physical trend that "dew point increases logarithmically with capacitance and shifts linearly with ambient temperature," thus providing a thermodynamically based benchmark anchor for subsequent algorithms.

[0142] Step 4: The MCU chip U3 executes a physically guided residual learning algorithm to perform high-precision compensation for the sensor's inherent nonlinear hysteresis and temperature-humidity cross-sensitivity in low-humidity and sub-zero environments. This includes the following sub-steps:

[0143] Step 4.1: The MCU chip U3 uses the measured smoothing capacitance value... Ambient temperature T and the reference dew point temperature obtained in step 3 Constructing high-dimensional physical extended feature vectors in memory :

[0144]

[0145] Introducing a logarithmic term into this eigenvector and quadratic terms To physically characterize the second-order nonlinear response of polymer moisture-sensitive materials, a cross term is introduced. and Used to explicitly characterize temperature and humidity cross-coupling interference.

[0146] Step 4.2: The MCU chip U3 will transfer the feature vector The input is fed into a C-language logic tree deployed within the device. This logic tree is derived from a pre-trained lightweight XGBoost residual regression model. The lightweight XGBoost residual regression model is an optimization regression algorithm based on Extreme Gradient Boosting. Its core principle is to sequentially construct multiple base decision trees (usually CART regression trees). The training objective of each new tree is to fit the residual (or negative gradient) between the previous model's prediction and the true value, thereby gradually reducing the error. After forward inference across multiple decision trees and accumulation of leaf node weights, the output is the predicted residual value used for local precision correction. This model is specifically designed to fit the true dew point to a reference dew point temperature. Instead of directly fitting absolute dew point values, the small nonlinear residuals between the two dimensions are used, thus significantly reducing the computational complexity and overfitting risk of high-dimensional mapping.

[0147] Step 5: To ensure the absolute reliability of the pure data-driven model output and prevent "anti-physics" predictions caused by transient interference in the industrial field, the MCU chip U3 monitors the prediction residual values. After performing post-processing physical hard constraints, the final dew point temperature (denoted as ) is finally calculated. The specific constraint logic is as follows:

[0148] Step 5.1: Residual amplitude truncation. Determine the predicted residual value. Whether the deviation exceeds the sensor's maximum allowable nonlinear deviation threshold M (preferably configured as 5°C in this embodiment), and perform amplitude limiting processing to obtain the truncated residual (denoted as ). ):

[0149]

[0150] Step 5.2: Absolute Thermodynamic Boundary Constraints and Output. According to the definition in physics, the dew point temperature of a gas can never be higher than its current ambient dry-bulb temperature. The microcontroller 105 will output the reference dew point temperature. Cut-off residuals Combined with the ambient temperature T, the final dew point temperature is output using the following formula:

[0151]

[0152] Step 6: Data Output

[0153] The MCU chip U3 will calculate the final dew point temperature. After smoothing, the signal is converted into an analog current signal by the 4-20mA module 103, or encapsulated into a Modbus protocol data packet by the RS-485 module 104 and output to the outside for reading by the industrial control system.

[0154] experiment

[0155] To verify the actual effectiveness and advancement of the capacitive dew point sensor acquisition device and dew point measurement method described in this invention, a comparative experiment was designed and conducted. The core objective of this experiment is to demonstrate the effectiveness of the "three-phase differential measurement hardware and dew point measurement method," and to showcase the outstanding substantive features and significant progress of this technology across the entire measurement range (especially in the low-humidity nonlinear range) compared to existing industry standard algorithms and conventional machine learning algorithms.

[0156] In terms of experimental data composition and acquisition, this experiment constructed a precision dew point generation and calibration system based on the principle of split-flow mixing. This system uses a zero-stage air filter to purify a high-pressure gas source and then splits it into two paths: one path is processed by a membrane dryer to generate an extremely dry reference gas, and the other path serves as a moisture bypass. By dynamically adjusting the ratio of dry to wet airflow using high-precision needle valves and proportional valves, the system can stably generate a standard dew point environment ranging from -40℃ to 5℃ within a multi-channel test chamber. To ensure the absolute authenticity and reliability of the calibration data, the experiment used a DM9600 cold mirror precision dew point meter (with a measurement accuracy of ±0.15℃) with extremely high metrological traceability as a reference standard. Under different date and temperature / humidity combinations, the experiment independently collected approximately 26,000 samples to form the training set, and approximately 24,000 samples collected at completely independent time periods to form the test set. For each sample, the pure capacitance value extracted from the front end of the device, the ambient temperature, and the true dew point temperature of the cold mirror were recorded simultaneously, thus effectively ensuring the generalization verification capability of the algorithm model. In the experiment, the semi-empirical physical baseline model used in this invention pre-determined system calibration coefficients through least squares fitting, with a = 0.797419, b = 15.288154, and c = -55.091098. The lightweight XGBoost residual regression model has 100 trees, a maximum tree depth of max_depth=4, and a learning rate of learning_rate=0.1.

[0157] In selecting comparative techniques, the experiment chose two representative existing techniques as quantitative comparison benchmarks. The first is the traditional lookup table method, which is currently the most widely used method in industrial dew point meters. This method selects multiple discrete calibration points, uses piecewise linear interpolation to convert capacitance to relative humidity, and then uses a standard meteorological formula to invert the dew point temperature. The second is the standard SVR (Support Vector Machine Regression) algorithm, which uses a radial basis function kernel function to directly establish the mapping from capacitance and temperature to dew point temperature, representing a pure data-driven black-box model without physical constraints. The experiment input over 24,000 independent test set samples into the two existing techniques and the dew point measurement method of this invention, respectively, for comprehensive error quantification evaluation. The quantitative error comparison results are shown in Table 1 below: Table 1

[0158] Existing technology 1: Traditional table lookup method 1.8165℃ 1.8614℃ 3.2668℃ Existing technology 2: Standard SVR algorithm 0.4876℃ 0.8838℃ 3.6764℃ This invention technology 0.1217℃ 0.2142℃ 1.6843℃

[0159] Quantitative analysis of the experimental results shows that the traditional lookup table method has a mean absolute error (MAE) as high as 1.8165℃, with a maximum absolute error of 3.2668℃. Its main drawback is that simple linear interpolation cannot accurately fit the severe nonlinear drift of the sensor in the low-humidity range below -30℃, resulting in a significant positive systematic bias. On the other hand, although the standard SVR algorithm reduces the mean absolute error to 0.4876℃, the lack of thermodynamic constraints in the purely data-driven model leads to severe nonlinear overfitting oscillations in local complex feature ranges (such as around -20℃), causing its maximum absolute error to surge to 3.6764℃, posing a significant risk of outputting "anti-physics" outliers. In contrast, the dew point measurement method of this invention exhibits superior performance, with a mean absolute error of only 0.1217℃. Thanks to the combined effect of the physical baseline model and the post-processing hard boundary constraint mechanism, this invention successfully avoids the output of extreme outliers, strictly controlling the maximum absolute error within 1.6843℃.

[0160] In summary, the experimental results fully demonstrate that this invention not only completely eliminates the interference of parasitic noise at the hardware physical level, but also achieves a perfect fusion of "physical interpretability" and "high-precision data fitting" at the software algorithm level. When facing harsh industrial measurement and control environments with wide temperature and humidity ranges, this system successfully overcomes the local overfitting defects of conventional data-driven models, exhibiting measurement accuracy and robust stability far exceeding existing technologies, demonstrating outstanding advancement and extremely high engineering implementation value.

[0161] Finally, it should be noted that the above examples are merely some specific embodiments of the present invention. Obviously, the present invention is not limited to the above embodiments and many variations are possible. All variations that can be directly derived or conceived by those skilled in the art from the disclosure of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A capacitive dew point sensor acquisition device, characterized in that: It includes a protective cover (1), a sensor base (2) and a main body shell (7) connected in sequence. A sensor capacitor C1 is provided inside the protective cover (1), a conductive connector (4) is provided inside the sensor base (2), and a data acquisition circuit board (6) is provided inside the main body shell (7). The acquisition circuit board (6) includes a sensor connection pad (108), a humidity detection module (106), a temperature detection module (107), and a microcontroller (105). The two ends of the conductive connector (4) are electrically connected to the sensor connection pad (108) and the sensor capacitor C1, respectively. The humidity detection module (106) includes a sensing measurement channel and a reference measurement channel. The sensing measurement channel is connected to the sensor connection pad (108) and the microcontroller (105) via signals. The reference measurement channel is connected to the microcontroller (105) via signals. The temperature detection module (107) is connected to the microcontroller (105) via signals. The sensor connection pad (108) is located at the front end of the acquisition circuit board (6). The humidity detection module (106) and the temperature detection module (107) are located close to each other behind the sensor connection pad (108). The gap between the sensor base (2) and the conductive connector (4) is filled with polymer sealant.

2. The capacitive dew point sensor acquisition device according to claim 1, characterized in that: The sensor base (2) is connected to the protective cover (1) and the main body shell (7) at both ends by threads. There are sealing elements between the protective cover (1) and the sensor base (2) and between the sensor base (2) and the main body shell (7).

3. The capacitive dew point sensor acquisition device according to claim 2, characterized in that: The sensing and measurement channel includes a first analog switch matrix U1. The VCC power supply is connected to the power input terminal of the first analog switch matrix U1 through a charging resistor R1. The first electrode and the second electrode of the sensor capacitor C1 are respectively connected to the channel selection terminal of the first analog switch matrix U1. The reference measurement channel includes a second analog switch matrix U2 and an onboard reference capacitor. The VCC power supply is connected to the power input terminal of the second analog switch matrix U2 via charging resistor R3; the reference capacitor... The first and second electrodes are respectively connected to the channel selection terminals of the second analog switch matrix U2; The microcontroller (105) includes an MCU chip U3, which is connected to the first analog switch matrix U1 and the second analog switch matrix U2 via an SPI communication bus. The SPI communication bus includes a common clock signal line (CLK), a master-in-slave-out signal line (MISO), and a master-out-slave-in signal line (MOSI). The first chip select terminal of MCU chip U3 is connected to the synchronization signal terminal of the first analog switch matrix U1, and the first signal acquisition terminal is connected to the sensing measurement terminal of the first analog switch matrix U1 as a sensing measurement bus (MCU_COM). The second chip select terminal of MCU chip U3 is connected to the synchronization signal terminal of the second analog switch matrix U2, and the second signal acquisition terminal is connected to the reference measurement terminal of the second analog switch matrix U2 as a reference measurement bus (MCU_COM1). The temperature detection module includes a voltage divider network, which is composed of an NTC thermistor and a precision reference resistor connected in series. The VCC power supply is grounded after passing through the voltage divider network. The common terminal of the NTC thermistor and the precision reference resistor is connected to the analog-to-digital conversion input channel of the MCU chip U3.

4. A method for measuring dew point using a capacitive dew point sensor acquisition device as described in any one of claims 1-3, characterized in that... The specific steps are as follows: Step 1: The microcontroller (105) constructs an analog capacitor detection circuit through the first analog switch matrix U1 and the second analog switch matrix U2 respectively; Then, the sensor capacitance C1 is continuously sampled. During each sampling, the sensing measurement channel is selected first, and multi-phase charge-discharge measurement and differential operation are performed on the sensor capacitance C1 to obtain the net discharge time T. net Then select the reference measurement channel for the onboard reference capacitor. Perform multi-phase charge-discharge measurements and differential calculations to obtain the reference time. Then, the absolute capacitance value of the sensor is obtained by using the ratio measurement principle. ; Step 2: Set the absolute capacitance value of the sensor Filtering is performed to obtain the smoothed capacitance value. ; Step 3: Synchronously collect the ambient temperature T, and calculate the reference dew point temperature using a preset physical baseline model. ; Step 4: Based on smoothing capacitance value Ambient temperature T and reference dew point temperature Constructing Physically Extended Eigenvectors Physically extended feature vectors The input is fed into a pre-trained residual regression model, and the output is the predicted residual value. ; Step 5: Predict the residual values Perform residual amplitude truncation to obtain truncated residuals ; Set the reference dew point temperature With truncated residuals Combined, the final dew point temperature is calculated. .

5. The method for measuring dew point according to claim 4, characterized in that: The multi-phase charge-discharge measurement process includes: First phase measurement: Construct a measurement circuit with sensor capacitance C1 and parasitic capacitance Cp1 in parallel and charge and discharge it, and record the first discharge time T1; Second phase measurement: Construct a measurement circuit with sensor capacitance C1 and parasitic capacitance Cp2 in parallel and charge and discharge it, and record the second discharge time T2; Third phase measurement: The two ends of the sensor capacitor C1 are shielded at the same potential, and a measurement circuit containing only the parasitic capacitances Cp1 and Cp2 in parallel is constructed and charged and discharged. The third discharge time T3 is recorded. The formula for calculating the difference in net discharge time is: .

6. The method for measuring dew point according to claim 5, characterized in that: The absolute capacitance value of the sensor The calculation formula is: , in, For the onboard reference capacitor The reference time is obtained by performing the multi-phase charge-discharge measurement and the differential calculation. The smoothing capacitance value The calculation formula is: ,in, The absolute capacitance value of the sensor is obtained in the nth acquisition, and N is the preset capacity of the static capacitor array.

7. The method for measuring dew point according to claim 6, characterized in that: The physical baseline model is as follows: ; Where C0 is the zero-point capacitance background value of the sensor under absolutely dry conditions; parameters a, b, and c are system calibration coefficients determined in advance by least squares fitting. The physical extended feature vector for: 。 8. The method for measuring dew point according to claim 7, characterized in that: The formula for the residual amplitude truncation process is expressed as follows: Where M is the maximum nonlinear deviation threshold; The formula for calculating the final dew point temperature is as follows: 。 9. The method for measuring dew point according to claim 8, characterized in that: The analog capacitor detection circuit constructed through the first analog switch matrix U1 includes gating switches A0 and A1, grounding switches B0 and B1, charging switch D1, and discharging switch D0. The VCC power supply terminal is connected to the sensor measurement bus via a charging resistor and a charging switch D1. The sensor measurement bus is connected to ground (GND) via a discharge switch D0 and a discharge resistor. The first electrode of sensor capacitor C1 is connected to one end of selector switch A1 and one end of ground switch B1, respectively. The other end of selector switch A1 is connected to the sensor measurement bus, and the other end of ground switch B1 is connected to ground (GND). The second electrode of sensor capacitor C1 is connected to one end of selector switch A0 and one end of ground switch B0, respectively. The other end of selector switch A0 is connected to the sensor measurement bus, and the other end of ground switch B0 is connected to ground (GND). In the measurement state, there is an equivalent parasitic capacitance Cp1 between the first electrode of sensor capacitor C1 and ground (GND), and an equivalent parasitic capacitance Cp2 between the second electrode of sensor capacitor C1 and ground (GND). The microcontroller connects sensor capacitor C1, parasitic capacitance Cp1, and parasitic capacitance Cp2 to the sensing measurement bus and performs charging and discharging operations by combining and controlling the switching on and off of gate switches A0 and A1, ground switches B0 and B1, and charging switch D1 and discharging switch D0.

Images