Temperature sensing system and method

Abstract

The invention relates to the technical field of temperature detection, in particular to a temperature sensing system and a temperature measuring method. A buffer amplifier is connected with an MEMS resonator through an interface circuit to generate a resonant frequency signal at the current moment. The time-to-digital converter is deployed in the field-programmable gate array and is used for synchronously acquiring the period number of the reference clock and the time deviation between the rising edge of the resonant frequency signal at the current moment after frequency division and the rising edge of the reference clock so as to obtain the corrected temperature; the temperature compensation unit is used for acquiring an initial correction temperature based on the initial temperature at the current moment, a mapping model between the temperature and the period number of a reference clock, and a mapping model between the temperature and the rising edge of a resonant frequency signal after frequency division processing and the time deviation of the rising edge of the reference clock; and carrying out overall calibration on the time-to-digital converter by using an iterative compensation algorithm. According to the invention, the influence of temperature drift is reduced, and the stability and reliability of the temperature sensing system are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of temperature detection, and in particular to a temperature sensing system and method. Background Art

[0002] As a fundamental measurement technology, temperature sensing plays a vital role in numerous fields, including industrial control, energy systems, intelligent manufacturing, healthcare, environmental monitoring, aerospace, and consumer electronics. In power systems, accurate temperature monitoring is crucial for preventing equipment overheating, improving energy efficiency, and extending service life. In the medical field, precise monitoring of human core temperature is directly linked to vital sign assessment and emergency response. In environmental monitoring, continuous temperature monitoring is fundamental to climate modeling and pollution analysis. And in consumer electronics, from smartphones and battery management systems to wearable devices, high-performance temperature sensors are increasingly demanding smaller sizes, lower power consumption, and faster response times.

[0003] The continuous development of emerging technologies such as the Internet of Things (IoT), artificial intelligence (AI), and edge computing has placed higher demands on intelligent, integrated, and high-performance temperature sensors. Traditional temperature sensors, such as thermistors and resistance temperature detectors (RTDs), while offering advantages in maturity and cost, have limitations in miniaturization, high response speed, long-term stability, and system integration. These limitations make them unable to fully meet the high-resolution, low-power, and embedded deployment requirements of next-generation intelligent systems. Therefore, promoting the development of novel temperature sensing technologies has become a key issue urgently needed in intelligent sensing systems. In recent years, temperature sensing solutions based on microelectromechanical system (MEMS) resonators have attracted significant attention due to their unique frequency-temperature characteristics. This technology utilizes temperature-induced resonant frequency shifts. As the temperature increases or decreases, the resonant frequency of the MEMS resonator shifts accordingly. By measuring this frequency shift and basing the functional relationship between temperature and frequency, the temperature of the measured environment can be determined. These solutions offer the advantages of high Q and sub-millisecond response.

[0004] However, MEMS resonator-based temperature sensors often use a spectrum analyzer to measure the resonant frequency of the MEMS resonator. While spectrum analyzers can accurately measure resonant frequency, they are bulky, power-hungry, expensive, and require an external power supply and complex operating interfaces, hindering the miniaturization and integration of the entire temperature sensing system. In applications such as the Internet of Things, wearable devices, and implantable medical monitoring, which place stringent requirements on sensor size, power consumption, and integration, this spectrum analyzer-based solution struggles to meet deployment requirements, severely limiting the promotion and application of MEMS resonator temperature sensing technology. Summary of the Invention

[0005] To this end, the technical problem to be solved by the present invention is to overcome the defects of existing MEMS resonator-based temperature sensors that use a spectrum analyzer to measure the resonant frequency of the MEMS resonator, making it difficult to achieve system miniaturization and integrated deployment, and unable to meet the requirements of low power consumption, miniaturization and high integration in scenarios such as the Internet of Things and wearable devices.

[0006] To solve the above technical problems, the present invention provides a temperature sensing system, comprising: A MEMS oscillator is used to generate a resonant frequency signal at the current moment; The time-to-digital converter, deployed in a field programmable gate array, is connected to the MEMS oscillator and receives the current resonant frequency signal, including: Temperature compensation unit, including: The initial temperature acquisition subunit is used to obtain the initial temperature at the current moment based on the number of cycles of the reference clock at the current moment collected by the coarse quantization module and the time deviation between the rising edge of the resonant frequency signal at the current moment output by the decoder after frequency division processing and the rising edge of the reference clock; A clock compensation subunit, configured to obtain a predicted number of cycles of the reference clock at the current moment based on the initial temperature at the current moment and a mapping model between the temperature and the number of cycles of the reference clock; a delay chain compensation subunit, configured to obtain a predicted time deviation between the rising edge of the resonant frequency signal after frequency division processing and the rising edge of the reference clock at the current moment based on a mapping model between the initial temperature at the current moment, the temperature, and the time deviation between the rising edge of the resonant frequency signal after frequency division processing and the rising edge of the reference clock; The initial temperature correction subunit is used to obtain the initial correction temperature at the current moment based on the predicted number of cycles of the reference clock at the current moment and the predicted time deviation between the rising edge of the resonant frequency signal after frequency division processing and the rising edge of the reference clock; The target temperature acquisition subunit is used to return the current moment initial correction temperature as the current moment initial temperature of the next iteration to the clock compensation subunit, generate the current moment initial correction temperature of the next iteration, until the difference between the current moment initial correction temperature and the initial temperature obtained in multiple consecutive iterations is less than the preset threshold, and the current moment initial correction temperature obtained in the last iteration is used as the current moment target temperature.

[0007] Preferably, the method of obtaining the initial temperature at the current moment based on the number of cycles of the reference clock at the current moment collected by the coarse quantization module and the time deviation between the rising edge of the resonant frequency signal at the current moment output by the decoder after frequency division processing and the rising edge of the reference clock includes: Calculate the initial period of the resonant frequency signal at the current moment based on the number of periods of the reference clock at the current moment collected by the coarse quantization module and the time deviation between the rising edge of the resonant frequency signal at the current moment output by the decoder after frequency division and the rising edge of the reference clock; Based on the initial period of the resonant frequency signal at the current moment, obtaining the initial frequency of the resonant frequency signal at the current moment; The initial temperature at the current moment is acquired based on the initial frequency of the resonant frequency signal at the current moment and a mapping model between the frequency of the resonant frequency signal and the temperature.

[0008] Preferably, the initial period of the resonant frequency signal at the current moment is calculated based on the number of periods of the reference clock at the current moment collected by the coarse quantization module and the time deviation between the rising edge of the resonant frequency signal at the current moment output by the decoder after frequency division processing and the rising edge of the reference clock, and the formula is: , in, is the frequency division coefficient, is the initial period of the resonant frequency signal at the current moment, is the number of cycles of the reference clock at the current moment, is the reference clock period, is the time deviation between the rising edge of the resonant frequency signal after frequency division and the rising edge of the reference clock at the current moment, is the rising edge time of the resonant frequency signal after frequency division at the current moment, The time of the rising edge of the reference clock at the current moment.

[0009] Preferably, the time-to-digital converter comprises: A pattern clock manager, whose output end is respectively connected to the coarse quantization module, the tapped delay chain and the input end of the temperature compensation unit for generating a reference clock; A frequency divider, whose input end is connected to the output end of the MEMS oscillator, is used to perform frequency division processing on the resonant frequency signal at the current moment; A coarse quantization module, whose input end is connected to the output end of the frequency divider, and whose output end is connected to the temperature compensation unit, is used to synchronously collect the number of cycles of the reference clock at the current moment when receiving the resonant frequency signal at the current moment, and transmit the number of cycles of the reference clock at the current moment to the temperature compensation unit; a tapped delay chain, the input of which is connected to the output of the frequency divider, for generating a state signal representing the time deviation between the rising edge of the resonant frequency signal at the current moment after the frequency division process and the rising edge of the reference clock; A decoder, whose input end is connected to the output end of the tapped delay chain, is used to convert the state signal output by the tapped delay chain into a binary code to obtain the time deviation between the rising edge of the resonant frequency signal at the current moment and the rising edge of the reference clock after frequency division processing; The storage module has an input end connected to the output end of the decoder and an output end connected to the temperature compensation unit. It is used to store the time deviation between the rising edge of the current resonant frequency signal output by the decoder after frequency division and the rising edge of the reference clock, and transmit it to the temperature compensation unit.

[0010] Preferably, the decoder is a type A encoder.

[0011] Preferably, the time-to-digital converter adopts a single tapped delay chain, which is composed of a plurality of delay units connected in cascade.

[0012] Preferably, the MEMS oscillator comprises: MEMS resonators, used to generate mechanical vibrations; An interface circuit, whose input terminal is connected to the output terminal of the MEMS resonator, is used to convert the mechanical vibration into an electrical signal to obtain the initial resonant frequency signal at the current moment; The buffer amplifier has an input end connected to the output end of the interface circuit, buffers and amplifies the initial resonant frequency signal at the current moment, and converts the level to meet the field programmable gate array standard to obtain the resonant frequency signal at the current moment.

[0013] Preferably, the process of acquiring the mapping model between temperature and the number of cycles of the reference clock includes: Through the hot box experiment, the number of reference clock cycles corresponding to different temperatures is obtained; Based on the number of reference clock cycles corresponding to different temperatures, a mapping model between temperature and the number of reference clock cycles is constructed.

[0014] Preferably, the process of acquiring the mapping model between the temperature and the time deviation between the rising edge of the resonant frequency signal after frequency division and the rising edge of the reference clock includes: Through the hot box experiment, the time deviation between the rising edge of the resonant frequency signal after frequency division processing and the rising edge of the reference clock corresponding to different temperatures is obtained; Based on the time deviation between the rising edge of the resonant frequency signal after frequency division processing and the rising edge of the reference clock corresponding to different temperatures, a mapping model between temperature and the time deviation between the rising edge of the resonant frequency signal after frequency division processing and the rising edge of the reference clock is constructed.

[0015] The present invention also provides a temperature measurement method, which is applied to the above-mentioned temperature sensing system, comprising: A MEMS resonator is used to generate a resonant frequency signal at the current moment and transmit it to a time-to-digital converter deployed in a field programmable gate array; The initial temperature at the current moment is obtained by the temperature compensation unit of the time-to-digital converter based on the number of cycles of the current reference clock collected by the coarse quantization module and the time deviation between the rising edge of the resonant frequency signal output by the decoder after frequency division and the rising edge of the reference clock. Obtaining a predicted number of cycles of the reference clock at the current moment based on a mapping model between the initial temperature at the current moment and the temperature and the number of cycles of the reference clock; Obtaining a predicted time deviation between the rising edge of the resonant frequency signal after frequency division processing and the rising edge of the reference clock at the current moment based on a mapping model between the initial temperature at the current moment, the temperature, and the time deviation between the rising edge of the resonant frequency signal after frequency division processing and the rising edge of the reference clock; Obtaining the initial correction temperature at the current moment based on the predicted number of cycles of the reference clock at the current moment and the predicted time deviation between the rising edge of the resonant frequency signal after frequency division and the rising edge of the reference clock; The initial correction temperature at the current moment is used as the initial temperature at the current moment of the next iteration, and the initial correction temperature at the current moment of the next iteration is generated and returned to execute the operation of the clock compensation subunit until the difference between the initial correction temperature at the current moment and the initial temperature obtained in multiple consecutive iterations is less than the preset threshold. The initial correction temperature at the current moment obtained in the last iteration is used as the corrected temperature at the current moment.

[0016] The above technical solution of the present invention has the following beneficial effects compared with the prior art: The temperature sensing system and temperature measurement method described in the present invention utilize a time-to-digital converter (TDC) deployed in a field programmable gate array (FPGA) and, in conjunction with a MEMS oscillator, can directly measure the frequency of a resonant frequency signal at the current moment. Considering that directly incorporating the TDC is susceptible to temperature influences, resulting in errors in the frequency measurement of the resonant frequency signal at the current moment, and thus inadequate temperature accuracy, the present invention incorporates a temperature compensation unit that obtains an initial temperature based on the number of reference clock cycles and the time deviation between the rising edge of the resonant frequency signal after frequency division and the rising edge of the reference clock. The unit also obtains an initial correction temperature by establishing mapping models between the temperature and the number of reference clock cycles, and between the temperature and the time deviation between the rising edge of the resonant frequency signal after frequency division and the rising edge of the reference clock. Using these two mapping models, the temperature compensation unit can correct the initial temperature to obtain the initial correction temperature. To further improve accuracy, multiple rounds of iterative correction are performed. Each round is based on the correction results of the previous round and is re-substituted into the model calculation to continuously reduce the error, and finally obtain a temperature value that accurately reflects the current moment. While realizing the miniaturization and integration of the system, the accuracy and stability of the final temperature reading are ensured to meet the application requirements of scenarios such as the Internet of Things and wearable devices. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to make the content of the present invention more clearly understood, the present invention is further described in detail below based on specific embodiments of the present invention in conjunction with the accompanying drawings, wherein: Figure 1 It is a structural diagram of a temperature sensing system of the present invention.

[0018] Figure 2 This is a flow chart of temperature correction processing of a temperature sensing system of the present invention.

[0019] Figure 3 Schematic diagram of the probability of bubbles forming in the delay chain at different temperatures.

[0020] Figure 4 This is the principle diagram of time interpolation measurement.

[0021] Figure 5 This is a schematic diagram of the MEMS oscillator output frequency at different temperatures.

[0022] Figure 6 It is a schematic diagram of the overall temperature test comparison.

[0023] Figure 7 This is a comparison diagram of the case where the temperature compensation unit is not introduced and the case where the temperature compensation unit is introduced. Figure 7 (a) is the measurement accuracy of the time-to-digital converter without the introduction of a temperature compensation unit. Figure 7 (b) in the figure is the measurement accuracy of the time-to-digital converter with the temperature compensation unit introduced. DETAILED DESCRIPTION

[0024] The present invention will be further described below with reference to the accompanying drawings and specific embodiments so that those skilled in the art can better understand the present invention and implement it. However, the embodiments are not intended to limit the present invention.

[0025] Reference Figure 1 As shown, this embodiment provides a temperature sensing system, including: A MEMS oscillator is used to generate a resonant frequency signal at the current moment; The time-to-digital converter is deployed in a field programmable gate array and connected to the MEMS oscillator. When receiving the current resonant frequency signal, it synchronously collects the number of cycles of the current reference clock and the time deviation between the rising edge of the resonant frequency signal after frequency division and the rising edge of the reference clock to obtain the corrected temperature at the current moment. The converter includes: like Figure 2 As shown, Figure 2 This is a flow chart of temperature correction processing of a temperature sensing system of the present invention.

[0026] Temperature compensation unit, including: The initial temperature acquisition subunit is used to obtain the initial temperature at the current moment based on the number of cycles of the reference clock at the current moment collected by the coarse quantization module and the time deviation between the rising edge of the resonant frequency signal at the current moment output by the decoder after frequency division processing and the rising edge of the reference clock; In this embodiment, specifically, obtaining the initial temperature at the current moment based on the number of cycles of the reference clock at the current moment collected by the coarse quantization module and the time deviation between the rising edge of the resonant frequency signal at the current moment output by the decoder after frequency division processing and the rising edge of the reference clock includes: Based on the number of cycles of the current reference clock collected by the coarse quantization module and the time deviation between the rising edge of the current resonant frequency signal output by the decoder after frequency division and the rising edge of the reference clock, the initial period of the current resonant frequency signal is calculated. The calculation formula is: , in, is the frequency division coefficient, used to reduce the measured frequency. is the initial period of the resonant frequency signal at the current moment, The number of cycles of the reference clock at the current moment recorded by the coarse quantization module, is the reference clock period, It is the time deviation between the rising edge of the resonant frequency signal after frequency division and the rising edge of the reference clock at the current moment, that is, the fine-grained delay compensation time, corresponding to the delay between Start and Stop. is the rising edge time of the resonant frequency signal after frequency division at the current moment, The time of the rising edge of the reference clock at the current moment.

[0027] Based on the initial period of the resonant frequency signal at the current moment, obtaining the initial frequency of the resonant frequency signal at the current moment; In this embodiment, specifically, the basic physical relationship between frequency and period is utilized, and the initial frequency of the resonant frequency signal at the current moment is the inverse of the initial period of the resonant frequency signal at the current moment to obtain the initial frequency of the resonant frequency signal at the current moment.

[0028] The initial temperature at the current moment is obtained based on the initial frequency of the resonant frequency signal at the current moment and a mapping model (P-MEMS) between the frequency of the resonant frequency signal and the temperature.

[0029] A clock compensation subunit, configured to obtain a predicted number of cycles of the reference clock at the current moment based on the initial temperature at the current moment and a mapping model between the temperature and the number of cycles of the reference clock; In this embodiment, specifically, the process of acquiring the mapping model between temperature and the number of cycles of the reference clock includes: During the factory delivery phase, a hot box test is performed to obtain the number of reference clock cycles corresponding to different temperatures. Based on the number of reference clock cycles corresponding to different temperatures, a mapping model between temperature and the number of reference clock cycles is constructed as a reference clock drift model (p-clk).

[0030] a delay chain compensation subunit, configured to obtain a predicted time deviation between the rising edge of the resonant frequency signal after frequency division processing and the rising edge of the reference clock at the current moment based on a mapping model between the initial temperature at the current moment, the temperature, and the time deviation between the rising edge of the resonant frequency signal after frequency division processing and the rising edge of the reference clock; In this embodiment, specifically, the process of acquiring the mapping model between the temperature and the time deviation between the rising edge of the resonant frequency signal after frequency division and the rising edge of the reference clock includes: During the factory delivery phase, a hot box test is performed to obtain the time deviation between the rising edge of the resonant frequency signal after frequency division and the rising edge of the reference clock corresponding to different temperatures. Based on the time deviation between the rising edge of the resonant frequency signal after frequency division and the rising edge of the reference clock corresponding to different temperatures, a mapping model between the temperature and the time deviation between the rising edge of the resonant frequency signal after frequency division and the rising edge of the reference clock is constructed as a delay chain drift model (p-TDL).

[0031] The initial temperature correction subunit is used to obtain the initial correction temperature at the current moment based on the predicted number of cycles of the reference clock at the current moment and the predicted time deviation between the rising edge of the resonant frequency signal after frequency division processing and the rising edge of the reference clock; The target temperature acquisition subunit is used to return the current moment initial correction temperature as the current moment initial temperature of the next iteration to the clock compensation subunit, generate the current moment initial correction temperature of the next iteration, until the difference between the current moment initial correction temperature and the initial temperature obtained in multiple consecutive iterations is less than the preset threshold, and the current moment initial correction temperature obtained in the last iteration is used as the current moment target temperature.

[0032] In this embodiment, specifically, the initial correction temperature at the current moment is returned to the clock compensation subunit as the initial temperature at the current moment of the next iteration, and the initial correction temperature at the current moment of the next iteration is generated, until the difference between the initial correction temperature at the current moment and the initial temperature obtained in two consecutive iterations is less than the preset threshold value, and the initial correction temperature at the current moment obtained in the last iteration is used as the target temperature at the current moment.

[0033] When the difference between the results of two consecutive iterations is lower than a preset threshold, the temperature estimation is considered to have converged. Actual tests have shown that this process usually only requires a small number of iterations to complete, with extremely high computational efficiency and convergence speed. In addition, because this solution is completely derived based on frequency data, it can achieve autonomous temperature compensation without the need for external temperature sensors. Compared with traditional static or peripheral-dependent methods such as lookup tables and XADC external temperature acquisition methods, it has higher system consistency and integrability, as well as excellent temperature adaptability and real-time performance. The system can track and compensate for temperature drift effects in real time, maintaining measurement accuracy within a wide temperature range of -40°C to 80°C.

[0034] In this embodiment, specifically, the MEMS oscillator includes: MEMS resonators, used to generate mechanical vibrations; An interface circuit, whose input terminal is connected to the output terminal of the MEMS resonator, is used to convert the mechanical vibration into an electrical signal to obtain the initial resonant frequency signal at the current moment; The buffer amplifier has an input end connected to the output end of the interface circuit, buffers and amplifies the initial resonant frequency signal at the current moment, and converts the level to meet the field programmable gate array standard to obtain the resonant frequency signal at the current moment.

[0035] As a core temperature-sensitive component, the MEMS oscillator utilizes the temperature-dependent frequency variation of its internal MEMS resonator to convert ambient temperature information into a frequency output signal. This oscillator integrates an internal oscillation circuit and generates the measured frequency signal using an external DC power supply. The output signal is shaped by a buffer amplifier to conform to the LVCMOS33 level standard and can be directly connected to the FPGA system through the I / O port, eliminating the need for external oscillator startup or signal conditioning circuitry. Its structure is simple and easy to connect.

[0036] For a temperature sensor based on a MEMS resonator, its resolution can be calculated using the resonator's temperature coefficient of frequency (TCF, in ppm / °C). The temperature resolution can be expressed as: , in, is the temperature resolution, Frequency measurement resolution, is the reference frequency of the resonator, is the temperature coefficient of the MEMS resonator. The temperature resolution depends on the frequency resolution, so the system requires high frequency resolution.

[0037] Existing resonant sensor measurement circuits often use traditional counting principles, which cannot fundamentally overcome the ±1 count error. Therefore, they still suffer from insufficient precision in scenarios requiring high accuracy. Furthermore, when faced with temperature fluctuations, the drift of components in various parts of the measurement circuit can have a significant impact on the final measurement results. This invention, through structural optimization and the introduction of a calibration mechanism, effectively improves the temperature robustness and measurement accuracy of the time-to-digital converter, thereby further enhancing system accuracy.

[0038] A MEMS oscillator is constructed by combining a MEMS resonator with an interface circuit. An amplifier generates a signal that meets FPGA-level standards and inputs it into a time-to-digital converter (TDC). The TDC utilizes a delay chain structure consisting of multiple cascaded delay cells to sample the signal to be measured in parallel at multiple nodes, achieving sub-clock time interval digitization. The sampled signal is latched by a high-speed trigger circuit and encoded and output in subsequent logic. However, the propagation delay of each cell in the delay chain is affected by process variations and temperature changes, resulting in non-uniformity in the sampling width.

[0039] To improve timing resolution and measurement accuracy, the system performs delay chain nonlinearity calibration during initial deployment. Statistical methods, such as code density testing, are used to measure and record the effective width of each bit. A mapping table converts the nonlinear bits into equivalent uniform least significant bits (LSBs). This corrects for nonlinear errors in the time-to-digital converter output, ensuring consistent time measurement accuracy across the full operating temperature range.

[0040] The present invention provides a temperature sensing system with an embedded reference clock source and a MEMS oscillator, the frequencies of which are affected by temperature and change. By modeling the relationship between the frequency of the MEMS oscillator and the temperature change, combined with the frequency measurement data obtained by the time-to-digital converter, a prediction model for the reference clock drift (a mapping model between the temperature and the number of cycles of the reference clock) is established to provide basic data for subsequent temperature estimation. A one-time temperature estimation is performed based on the initial frequency measurement result, and then the temperature drift of the time-to-digital converter itself and the reference clock is reversed based on the estimated value to correct the frequency measurement result. The compensation process is performed in an iterative manner until the estimated temperature value converges, thereby achieving a closed-loop correction of self-calibration. During the calibration phase, the system obtains multiple sets of temperature-frequency drift parameters through hot box experiments, including MEMS resonator characteristic curves, delay chain temperature drift characteristics, reference clock drift curves, etc., to form a configurable temperature drift parameter set. During measurement, the corresponding parameters are dynamically loaded by the control logic to improve the compensation accuracy.

[0041] To overcome the shortcomings of existing MEMS resonator-based temperature sensors that use a spectrum analyzer to measure the resonant frequency of the MEMS resonator, making it difficult to achieve system miniaturization and integrated deployment, and failing to meet the low power consumption, miniaturization, and high integration requirements of scenarios such as the Internet of Things (IoT) and wearable devices, this invention introduces a time-to-digital converter (TDC) as a signal processor to measure the resonant frequency of the MEMS resonator. However, the TDC is susceptible to temperature effects, which can lead to errors in the temperature reading and reduced accuracy. To address the temperature deviation of the TDC, existing research has proposed a delay chain calibration scheme at different temperatures. This work measures the different delays of the TDC's delay chain at different temperatures and uses the XADC temperature sensor within the FPGA for temperature compensation, achieving a root mean square resolution of 5 ps. However, this approach relies on an external temperature sensor for temperature compensation. However, this reliance on an external temperature sensor increases the system's hardware complexity and power consumption, undermining the original design goals of miniaturization and integration. Furthermore, temperature response differences and spatial temperature gradients between the external sensor and the TDC can introduce additional compensation errors, making high-precision real-time calibration difficult. This makes it unsuitable for scenarios such as the IoT and wearable devices, which require stringent low power consumption, high integration, and measurement accuracy. Therefore, the present invention integrates a temperature compensation unit inside the time-to-digital converter.

[0042] By establishing a temperature drift lookup table and using an iterative compensation algorithm to calibrate the time-to-digital converter as a whole, the impact of temperature drift is reduced, the stability and reliability of the temperature measurement system are improved, and temperature compensation of the entire system is achieved without introducing an external temperature sensor.

[0043] In this embodiment, specifically, the time-to-digital converter includes: A pattern clock manager, whose output end is respectively connected to the coarse quantization module, the tapped delay chain and the input end of the temperature compensation unit for generating a reference clock; A frequency divider, whose input end is connected to the output end of the MEMS oscillator, is used to perform frequency division processing on the resonant frequency signal at the current moment; A coarse quantization module, whose input end is connected to the output end of the frequency divider, and whose output end is connected to the temperature compensation unit, is used to synchronously collect the number of cycles of the reference clock at the current moment when receiving the resonant frequency signal at the current moment, and transmit the number of cycles of the reference clock at the current moment to the temperature compensation unit; a tapped delay chain, the input of which is connected to the output of the frequency divider, for generating a state signal representing the time deviation between the rising edge of the resonant frequency signal at the current moment after the frequency division process and the rising edge of the reference clock; A decoder, whose input end is connected to the output end of the tapped delay chain, is used to convert the state signal output by the tapped delay chain into a binary code to obtain the time deviation between the rising edge of the resonant frequency signal at the current moment and the rising edge of the reference clock after frequency division processing; The storage module has an input end connected to the output end of the decoder and an output end connected to the temperature compensation unit. It is used to store the time deviation between the rising edge of the current resonant frequency signal output by the decoder after frequency division and the rising edge of the reference clock, and transmit it to the temperature compensation unit.

[0044] A time-to-digital converter (TDC) deployed on an FPGA platform accurately acquires the period of the MEMS oscillator's output signal, thereby inferring the ambient temperature. The TDC is implemented using the Xilinx Zynq-7020 chip, model XC7Z020CLG400-2, manufactured using a 28nm CMOS process. The TDC internally comprises a frequency divider, a tapped delay chain, a coarse quantization module, a thermal decoder, a storage module, and a temperature compensation unit. It provides high-resolution time interval measurement and is suitable for inverse frequency-temperature mapping. The TDC employs a combined coarse and fine quantization architecture and leverages native FPGA resources (CARRY4 delay cells and a D-flip-flop array) to achieve high-resolution time interval measurement. The system consists of three main components: a coarse quantization module, a fine quantization module (TDL), a decoding module, a buffer module, and the output of the compensation module. These modules work together to achieve high-precision frequency reading and temperature mapping. To enable the deployment of temperature sensing devices in different temperature zones, the system architecture supports the deployment of multiple sensor heads. MEMS oscillators are located in each temperature-controlled area, while the FPGA master controller is located in a centralized acquisition center. The output frequency of each sensor node is collected through independent channels, converted to a digital signal locally using a time-controlled digital converter (TDC), and ultimately outputs the temperature result. Fine quantization outputs a binary result after ones-counter encoding, which is cached along with the coarse quantization data in a FIFO buffer module to ensure timing integrity. This data is then input into the system's temperature calibration module, where it is compensated and converted using preloaded temperature model parameters. The corresponding temperature value or other application data is ultimately output via the serial (UART) interface.

[0045] To expand the measurement dynamic range of the time-to-digital converter (TDC), the present invention incorporates a coarse quantization counter in addition to the fine quantization module to count the number of reference clock cycles. The system divides the frequency of the oscillator signal under test, using the divided signal as the "Start" trigger signal and the reference clock as the "Stop" signal. Large-scale time intervals are then accumulated by accumulating cycles. The TDC architecture not only uses the rising edge of the divided signal as the Start signal, but also utilizes its falling edge, combined with the rising edge of the reference clock, as the Stop signal. This allows for finer-grained time compensation, improves edge positioning accuracy, and reduces the impact of the ±1-cycle error found in traditional solutions.

[0046] To capture the state of the delayed signal, the system connects the output of each delay chain stage to an array of D flip-flops, which are sampled using a reference clock, generating a series of hot codes. In theory, this hot code should be monotonic, such as "00001111." However, in practice, due to system clock skew and conflicts in setup and hold times, bubbles (discontinuous "1s" or "0s" such as "0010011") can easily occur, leading to encoding errors.

[0047] To this end, the present invention adopts a single-chain TDL structure, significantly reducing the non-monotonic distortion caused by inconsistent delays compared to traditional multi-TDL structures. Furthermore, a ones-counter encoder is introduced in the encoding phase, replacing the traditional one-hot-to-binary priority encoder. This one-hot conversion is performed by "counting ones," effectively suppressing the impact of bubbles on the final quantization result and improving system robustness.

[0048] In terms of stability in high temperature environments, the system design fully considers the impact of temperature changes on the delay chain characteristics and reference clock stability.

[0049] like Figure 3 As shown, Figure 3Schematic diagram of the probability of bubble generation within a delay chain at different temperatures. Experimental results show that the probability of bubble distortion in a time-to-digital converter increases from 31% to 49.4% as the temperature rises from 30°C to 80°C, showing a clear exponential growth trend. This bubble aggregation typically occurs near the edges of hot code transitions, forming bubble clusters that severely interfere with the subsequent encoding process. A delay chain consists of multiple cascaded delay cells, and the propagation delay of each cell fluctuates with temperature. As the temperature rises or falls, the electrical properties (such as carrier mobility) of the semiconductor material (such as silicon) that makes up the delay cell change, resulting in a shortened or extended cell delay. Furthermore, temperature variations exacerbate the inherent non-uniformity caused by process variations, further increasing delay variations between cells. When this delay fluctuation or variation exceeds a certain threshold, the signal can be sampled simultaneously at multiple nodes in the delay chain, potentially leading to trigger timing errors and the generation of bubbles.

[0050] In response to this problem, this embodiment has made targeted optimizations in the structural design. Preferably, the time-to-digital converter adopts a single tapped delay chain, which is composed of a plurality of delay units in cascade. Compared with the traditional multi-chain structure, the single-chain structure can fundamentally avoid the characteristic differences between multiple chains caused by process deviations and temperature drifts, reduce the additional interference caused by multi-path signal transmission, and make the propagation delay characteristics of the delay unit easier to uniformly control and calibrate. At the same time, in the encoding link, this embodiment preferably uses a type-1 encoder instead of a traditional thermal code encoder. The type-1 encoder has stronger anti-interference ability and fault tolerance, and can effectively identify and suppress bubble signals generated by high temperature in the delay chain - even in scenarios where bubbles are densely concentrated at high temperatures, the type-1 encoder can also sort out and correct chaotic sampling signals through optimized encoding logic to avoid encoding errors caused by bubbles. The synergistic effect of the single-chain structure and the type-I encoder not only greatly alleviates the problem of dense bubble concentration at high temperatures and reduces the interference of bubbles on signal sampling, but also significantly improves the linearity and decodability of time measurement results, ensuring that the time-to-digital converter can output stable and reliable encoding results within a wide temperature range, laying a solid foundation for subsequent temperature compensation and precise measurement.

[0051] The present invention effectively improves the measurement accuracy and system stability of a high-precision time measurement system over a wide temperature range by constructing a time-to-digital converter (TDC) based on a tapped delay chain (TDL) in a field programmable gate array (FPGA) and combining it with a self-calibrating temperature compensation module, thus achieving significant technical, economic and social benefits.

[0052] like Figure 4 As shown, Figure 4 This is the principle diagram of time interpolation measurement.

[0053] In terms of time quantization accuracy, the present invention uses a multi-level CARRY4 structure to construct a fine quantization delay chain, combined with the time interpolation method of the system clock. Figure 2 As shown in the figure, this architecture combines coarse and fine time-delayed measurement (TDC) capabilities, achieving picosecond-level time interval measurement. The CARRY4 module in each delay chain stage can be logically configured to implement cascaded propagation of fast carry signals, ensuring monotonic progression of the time signal through the delay path, effectively improving fine time resolution. Experiments demonstrate that this architecture maintains stable delay characteristics and high-speed sampling capabilities when deployed on a Xilinx Zynq7020 FPGA.

[0054] In addition, in terms of resistance to temperature drift and system adaptability, this system proposes a temperature compensation unit. By introducing a closed-loop iterative compensation mechanism based on frequency measurement, it achieves full system-level compensation for errors caused by temperature changes in the MEMS oscillator, reference clock, and TDC delay chain without the need for external temperature sensors, further improving the system's long-term stability and wide-temperature measurement accuracy.

[0055] In this embodiment, specifically, it further includes: a display connected to the time-to-digital converter, which is used to display the corrected temperature at the current moment.

[0056] The temperature sensing system proposed in this invention consists of a MEMS oscillator and a reference clock source as the sensing front end. The output frequencies of both have significant temperature dependence. The frequency output of the MEMS oscillator can be expressed as , which shows a monotonic shift with temperature changes; the frequency of the reference clock shows a nearly linear drift relationship, which can be modeled as Both signals are input to TDC for frequency measurement, and the raw time data obtained directly reflects the current temperature status. is the output frequency of the reference clock source, is the linear coefficient of the reference clock frequency changing with temperature, Indicates temperature, is the initial frequency of the reference clock source at the reference temperature, that is, the intercept term of the reference clock frequency model. To achieve high-precision temperature recovery, the present invention first pre-calibrates the system in a constant temperature chamber, collecting TDC output offset data at multiple temperature points to obtain a number of systematic compensation parameters. These parameters are pre-loaded into the system and serve as the basis for subsequent compensation processing. The raw frequency data collected by the TDC is then transmitted to the temperature compensation engine for multiple error compensation and frequency correction processes. The core of this compensation engine is an offset predictor, which calculates the reference clock frequency drift based on the current initial temperature. The entire compensation process employs an iterative closed-loop structure: an initial temperature estimate is derived based on the uncorrected frequency; this initial temperature is then used to infer the reference clock drift and delay chain error, reversely correcting the frequency value, and recalculating the initial corrected temperature. This process continues until the difference between the current initial corrected temperature and the initial temperature is less than a preset threshold.

[0057] Any MEMS resonator with a stable TCF can be used to implement high-precision temperature sensing. A MAX7381AXR106-T MEMS oscillator was used as the front-end temperature acquisition component. To verify the temperature sensitivity of the MEMS resonator employed in this invention, a systematic test of its frequency behavior with temperature was conducted using a constant temperature chamber. In the experimental setup, the MEMS oscillator was placed in an incubator, and the ambient temperature was gradually increased from –40°C to 80°C, maintaining stability at each temperature point to ensure the accuracy and repeatability of the test results. The output signal was monitored in real time using a Siglent SNA5051X spectrum analyzer for frequency changes.

[0058] This system uses MAX7381AXR106-T as the MEMS oscillator to output frequency information. The actual performance of the proposed temperature compensation unit and the overall system is verified through system testing, such as Figure 5 As shown, Figure 5It is a schematic diagram of the output frequency of the MEMS oscillator at different temperatures. The temperature of the MEMS oscillator was scanned in the range of -40℃ to 80℃, and the measured frequency change showed a good linear thermal response, indicating that it has high temperature sensitivity and is suitable for temperature sensing applications. The reference frequency signal was further simulated using a signal generator, and the TDC frequency measurement experiment was performed on the 10.05 MHz signal in the range of -40℃ to 80℃ with a step of 10℃. The results showed that the measurement resolution of the system was significantly improved after the addition of dynamic temperature calibration, and the RMS resolution of TDC was increased from about 35ps to 23ps. The resonator inside the MEMS oscillator showed a nearly linear frequency drift trend in the entire temperature change range, and the frequency increased steadily with the increase in temperature, showing a significant thermal response characteristic. According to the measured data, its thermal frequency drift coefficient (TCF) is about 2056.5 ppm / ℃, indicating that the resonator has a highly linear and repeatable frequency response under temperature changes, and can effectively reflect changes in ambient temperature. Finally, the complete system was placed in a temperature control box for temperature measurement experiments. Figure 6 As shown, Figure 6 Figure 2 shows a comparison diagram of the overall temperature test. The system temperature error measured over the 20°C to 70°C range averaged 0.019°C, with a maximum of no more than 0.06°C. The uncalibrated system (a temperature sensing system without a temperature compensation unit) exhibited an average error of 0.243°C and a maximum of 0.53°C. This demonstrates that the proposed temperature compensation method significantly improves the system's temperature measurement accuracy and stability over a wide temperature range.

[0059] To evaluate the accuracy of the proposed temperature sensing system, a comparative experiment was designed and conducted. A signal generator was used to generate a reference signal with a fixed frequency close to the output frequency of the MEMS oscillator. In this experiment, the frequency was set to 10.05 MHz to simulate the operating conditions in a real temperature sensing environment. During the test, the system was placed in a constant temperature environment ranging from -40°C to 80°C, with a step of 10°C, for a total of 12 temperature points. 200 samples were collected at each temperature point, for a total of 24,000 independent measurements to statistically evaluate the system's frequency measurement stability and resolution over the full temperature range.

[0060] The measurement results are as follows Figure 7 As shown, Figure 7 This is a comparison diagram of the case where the temperature compensation unit is not introduced and the case where the temperature compensation unit is introduced. Figure 7 (a) is the measurement accuracy of the time-to-digital converter without the introduction of a temperature compensation unit. Figure 7 (b) in the figure is the measurement accuracy of the time-to-digital converter with the temperature compensation unit introduced.

[0061] The distribution of period measurements before and after calibration is statistically displayed using histograms. In the uncalibrated state, the TDC measured an average period of 99.543 ns with a root mean square (RMS) resolution of 34.88 ps. After compensation using the temperature calibration unit, the average period was corrected to 99.498 ns, and the resolution was improved to 23.55 ps. This demonstrates the present invention's excellent measurement consistency and high resolution in complex temperature environments. In particular, the implementation of the self-calibration and compensation algorithm further enhances system performance, meeting the requirements of high-precision temperature sensing applications.

[0062] Under stable DC power supply conditions, a segmented temperature control experiment was carried out in the range of 20℃ to 70℃, with each temperature step being 5℃. The temperature was measured using the calibration and compensation method implemented in this system. Figure 6 As shown, within the core operating temperature range of 20°C to 50°C, the system demonstrates extremely high measurement accuracy, with the maximum temperature error consistently remaining below 0.01°C. After temperatures exceed 50°C, the error gradually increases, reaching a maximum of 0.06°C at 70°C. Statistical analysis across the entire range reveals an average measurement error of 0.019°C. This compares to an uncalibrated system with a maximum error of 0.53°C and an average error of 0.243°C. This demonstrates that the self-calibrating temperature compensation mechanism proposed in this invention can significantly improve measurement accuracy in practical applications and possesses excellent engineering value.

[0063] While there is still a slight upward trend in error at high temperatures, primarily due to the increased phase noise of the MEMS oscillator at high temperatures, stemming from its intrinsic thermal noise characteristics, the overall system still maintains good precision control, validating the superior performance of this invention in high-resolution, high-stability temperature measurement scenarios.

[0064] The second embodiment provides a temperature measurement method, which is applied to the above-mentioned temperature sensing system, including: A MEMS resonator is used to generate a resonant frequency signal at the current moment and transmit it to a time-to-digital converter deployed in a field programmable gate array; The initial temperature at the current moment is obtained by the temperature compensation unit of the time-to-digital converter based on the number of cycles of the current reference clock collected by the coarse quantization module and the time deviation between the rising edge of the resonant frequency signal output by the decoder after frequency division and the rising edge of the reference clock. Obtaining a predicted number of cycles of the reference clock at the current moment based on a mapping model between the initial temperature at the current moment and the temperature and the number of cycles of the reference clock; Obtaining a predicted time deviation between the rising edge of the resonant frequency signal after frequency division processing and the rising edge of the reference clock at the current moment based on a mapping model between the initial temperature at the current moment, the temperature, and the time deviation between the rising edge of the resonant frequency signal after frequency division processing and the rising edge of the reference clock; Obtaining the initial correction temperature at the current moment based on the predicted number of cycles of the reference clock at the current moment and the predicted time deviation between the rising edge of the resonant frequency signal after frequency division and the rising edge of the reference clock; The initial correction temperature at the current moment is used as the initial temperature at the current moment of the next iteration, and the initial correction temperature at the current moment of the next iteration is generated and returned to execute the operation of the clock compensation subunit until the difference between the initial correction temperature at the current moment and the initial temperature obtained in multiple consecutive iterations is less than the preset threshold. The initial correction temperature at the current moment obtained in the last iteration is used as the corrected temperature at the current moment.

[0065] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0066] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0067] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0068] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps for the function specified in one or more boxes.

[0069] Obviously, the above embodiments are merely examples for clarity of explanation and are not intended to limit the implementation methods. Those skilled in the art will appreciate that other variations or modifications can be made based on the above description. It is not necessary and impossible to enumerate all implementation methods here. Obvious variations or modifications arising therefrom remain within the scope of protection of the present invention.

Claims

1. A temperature sensing system, characterized in that: include: A MEMS oscillator is used to generate a resonant frequency signal at the current moment; The time-to-digital converter, deployed in a field programmable gate array, is connected to the MEMS oscillator and receives the current resonant frequency signal, including: Temperature compensation unit, including: The initial temperature acquisition subunit is used to obtain the initial temperature at the current moment based on the number of cycles of the reference clock at the current moment collected by the coarse quantization module and the time deviation between the rising edge of the resonant frequency signal at the current moment output by the decoder after frequency division processing and the rising edge of the reference clock; A clock compensation subunit, configured to obtain a predicted number of cycles of the reference clock at the current moment based on the initial temperature at the current moment and a mapping model between the temperature and the number of cycles of the reference clock; a delay chain compensation subunit, configured to obtain a predicted time deviation between the rising edge of the resonant frequency signal after frequency division processing and the rising edge of the reference clock at the current moment based on a mapping model between the initial temperature at the current moment, the temperature, and the time deviation between the rising edge of the resonant frequency signal after frequency division processing and the rising edge of the reference clock; The initial temperature correction subunit is used to obtain the initial correction temperature at the current moment based on the predicted number of cycles of the reference clock at the current moment and the predicted time deviation between the rising edge of the resonant frequency signal after frequency division processing and the rising edge of the reference clock; The target temperature acquisition subunit is used to return the current moment initial correction temperature as the current moment initial temperature of the next iteration to the clock compensation subunit, generate the current moment initial correction temperature of the next iteration, until the difference between the current moment initial correction temperature and the initial temperature obtained in multiple consecutive iterations is less than the preset threshold, and the current moment initial correction temperature obtained in the last iteration is used as the current moment target temperature.

2. A temperature sensing system according to claim 1, characterized in that: The method of obtaining the initial temperature at the current moment based on the number of cycles of the reference clock at the current moment collected by the coarse quantization module and the time deviation between the rising edge of the resonant frequency signal at the current moment output by the decoder after frequency division processing and the rising edge of the reference clock includes: Calculate the initial period of the resonant frequency signal at the current moment based on the number of periods of the reference clock at the current moment collected by the coarse quantization module and the time deviation between the rising edge of the resonant frequency signal at the current moment output by the decoder after frequency division and the rising edge of the reference clock; Based on the initial period of the resonant frequency signal at the current moment, obtaining the initial frequency of the resonant frequency signal at the current moment; The initial temperature at the current moment is acquired based on the initial frequency of the resonant frequency signal at the current moment and a mapping model between the frequency of the resonant frequency signal and the temperature.

3. A temperature sensing system according to claim 2, characterized in that: The initial period of the resonant frequency signal at the current moment is calculated based on the number of periods of the reference clock at the current moment collected by the coarse quantization module and the time deviation between the rising edge of the resonant frequency signal at the current moment output by the decoder after frequency division processing and the rising edge of the reference clock. The formula is: ， in, is the frequency division coefficient, is the initial period of the resonant frequency signal at the current moment, is the number of cycles of the reference clock at the current moment, is the reference clock period, is the time deviation between the rising edge of the resonant frequency signal after frequency division and the rising edge of the reference clock at the current moment, is the rising edge time of the resonant frequency signal after frequency division at the current moment, The time of the rising edge of the reference clock at the current moment.

4. A temperature sensing system according to claim 1, characterized in that: The time-to-digital converter includes: A pattern clock manager, whose output end is respectively connected to the coarse quantization module, the tapped delay chain and the input end of the temperature compensation unit for generating a reference clock; A frequency divider, whose input end is connected to the output end of the MEMS oscillator, is used to perform frequency division processing on the resonant frequency signal at the current moment; A coarse quantization module, whose input end is connected to the output end of the frequency divider, and whose output end is connected to the temperature compensation unit, is used to synchronously collect the number of cycles of the reference clock at the current moment when receiving the resonant frequency signal at the current moment, and transmit the number of cycles of the reference clock at the current moment to the temperature compensation unit; a tapped delay chain, the input of which is connected to the output of the frequency divider, for generating a state signal representing the time deviation between the rising edge of the resonant frequency signal at the current moment after the frequency division process and the rising edge of the reference clock; A decoder, whose input end is connected to the output end of the tapped delay chain, is used to convert the state signal output by the tapped delay chain into a binary code to obtain the time deviation between the rising edge of the resonant frequency signal at the current moment and the rising edge of the reference clock after frequency division processing; The storage module has an input end connected to the output end of the decoder and an output end connected to the temperature compensation unit. It is used to store the time deviation between the rising edge of the current resonant frequency signal output by the decoder after frequency division and the rising edge of the reference clock, and transmit it to the temperature compensation unit.

5. A temperature sensing system according to claim 4, characterized in that: A decoder is a type of encoder.

6. A temperature sensing system according to claim 4, characterized in that: The time-to-digital converter adopts a single tapped delay chain, which is composed of a plurality of delay units cascaded.

7. A temperature sensing system according to claim 1, characterized in that: MEMS oscillators include: MEMS resonators, used to generate mechanical vibrations; An interface circuit, whose input terminal is connected to the output terminal of the MEMS resonator, is used to convert the mechanical vibration into an electrical signal to obtain the initial resonant frequency signal at the current moment; The buffer amplifier has an input end connected to the output end of the interface circuit, buffers and amplifies the initial resonant frequency signal at the current moment, and converts the level to meet the field programmable gate array standard to obtain the resonant frequency signal at the current moment.

8. A temperature sensing system according to claim 1, characterized in that: The process of obtaining the mapping model between temperature and the number of cycles of the reference clock includes: Through the hot box experiment, the number of reference clock cycles corresponding to different temperatures is obtained; Based on the number of reference clock cycles corresponding to different temperatures, a mapping model between temperature and the number of reference clock cycles is constructed.

9. A temperature sensing system according to claim 1, characterized in that: The process of obtaining the mapping model between the time deviation between the rising edge of the temperature and resonant frequency signal after frequency division and the rising edge of the reference clock includes: Through the hot box experiment, the time deviation between the rising edge of the resonant frequency signal after frequency division processing and the rising edge of the reference clock corresponding to different temperatures is obtained; Based on the time deviation between the rising edge of the resonant frequency signal after frequency division processing and the rising edge of the reference clock corresponding to different temperatures, a mapping model between temperature and the time deviation between the rising edge of the resonant frequency signal after frequency division processing and the rising edge of the reference clock is constructed.

10. A temperature measurement method, characterized in that: A temperature sensing system according to any one of claims 1 to 9, comprising: A MEMS resonator is used to generate a resonant frequency signal at the current moment and transmit it to a time-to-digital converter deployed in a field programmable gate array; The initial temperature at the current moment is obtained by the temperature compensation unit of the time-to-digital converter based on the number of cycles of the current reference clock collected by the coarse quantization module and the time deviation between the rising edge of the resonant frequency signal output by the decoder after frequency division and the rising edge of the reference clock. Obtaining a predicted number of cycles of the reference clock at the current moment based on a mapping model between the initial temperature at the current moment and the temperature and the number of cycles of the reference clock; Obtaining a predicted time deviation between the rising edge of the resonant frequency signal after frequency division processing and the rising edge of the reference clock at the current moment based on a mapping model between the initial temperature at the current moment, the temperature, and the time deviation between the rising edge of the resonant frequency signal after frequency division processing and the rising edge of the reference clock; Obtaining the initial correction temperature at the current moment based on the predicted number of cycles of the reference clock at the current moment and the predicted time deviation between the rising edge of the resonant frequency signal after frequency division and the rising edge of the reference clock; The initial correction temperature at the current moment is used as the initial temperature at the current moment of the next iteration, and the initial correction temperature at the current moment of the next iteration is generated and returned to execute the operation of the clock compensation subunit until the difference between the initial correction temperature at the current moment and the initial temperature obtained in multiple consecutive iterations is less than the preset threshold. The initial correction temperature at the current moment obtained in the last iteration is used as the corrected temperature at the current moment.

Images