A real-time digital voice intercom system based on industrial fieldbus

Abstract

This invention relates to the field of industrial automation communication technology, specifically a real-time digital voice intercom system based on an industrial fieldbus, comprising: a ground terminal device, a gateway device, and extension units. The ground terminal device collects and encodes voice signals, combines them with control commands, encapsulates them into Ethernet frames, and sends them to the gateway device. The gateway device performs bidirectional conversion between the industrial bus protocol and Ethernet, separates voice data from control commands, and employs a dynamic bandwidth allocation method based on voice activity detection and load prediction to optimize bus resource utilization and ensure low-latency transmission. The extension units receive voice signals to drive the output of explosion-proof speakers, collect signals from field devices, and support seamless roaming switching based on game theory power adjustment and channel prediction. This invention employs a ring topology redundancy design to achieve rapid fault self-healing, effectively adapting to high-noise explosion-proof industrial environments, and realizing reliable synchronous transmission of control commands and high-fidelity voice.

Description

Technical Field

[0001] This invention relates to the field of industrial automation communication technology, and more specifically to a real-time digital voice intercom system based on an industrial fieldbus. Background Technology

[0002] The demand for collaborative operation of voice communication and equipment control in the industrial sector is becoming increasingly significant. Industrial fieldbuses (such as RS-485, CAN, and Modbus) have been widely used as the core network for equipment control, while high-fidelity voice interaction is indispensable in scenarios such as dispatching and command, and safety early warning.

[0003] Chinese invention patent application CN112954038A discloses a converged communication system for a well site data acquisition device, including a converged communication gateway module, an audio / video acquisition module, a UHF walkie-talkie, and a remote command and dispatch platform. The audio / video acquisition module is used to acquire well site audio / video monitoring information and transmit it to the remote command and dispatch platform via the converged communication gateway module. The remote command and control platform is used to receive and display well site monitoring information via the converged communication gateway module and to conduct voice communication with the UHF walkie-talkie via the converged gateway communication module. This invention integrates well site data acquisition and transmission functions with voice communication functions into the converged communication gateway module. It uploads on-site environmental and audio / video monitoring information through the converged communication gateway and utilizes the networking function of the converged communication gateway module to achieve voice communication with the remote command and dispatch platform, facilitating platform personnel to monitor on-site information and conduct on-site command and dispatch simultaneously.

[0004] However, special industrial environments such as high noise and explosion-proof safety place clear demands on the adaptability of communication systems. There is an urgent need to develop an integrated system that combines industrial bus and voice communication to achieve synchronous transmission of control commands and voice data, and to meet the real-time, reliability and environmental adaptability requirements of industrial scenarios. Summary of the Invention

[0005] The purpose of this invention is to address the problems existing in the background technology by proposing a real-time digital voice intercom system based on industrial fieldbus.

[0006] The technical solution of this invention: A real-time digital voice intercom system based on an industrial fieldbus, comprising: Ground-based equipment is used to collect voice signals through a microphone array, encode them, encapsulate them with control commands into Ethernet frames, and send them to the gateway device. It also receives and restores the audio signals to drive the speaker output. Gateway devices are used to realize bidirectional conversion between industrial bus protocols and Ethernet, separate voice data and control commands, forward voice data to the target extension unit, convert control commands into industrial bus protocols and send them to the execution device, collect sensor data and send it back, and dynamically allocate bus resources using a time-division multiplexing mechanism. The extension unit is used to receive voice data to drive the output of the explosion-proof speaker, collect equipment switching signals and sensor signals, and supports wireless relay access and roaming switching.

[0007] Preferably, the ground-based equipment includes: Audio input / output unit, equipped with a noise-canceling microphone array and analog-to-digital converter and digital-to-analog converter; The processor unit integrates a DSP chip to perform encoding and decoding, and encapsulates data frames containing 8-byte control instruction fields; The network port PHY module provides 10, 100, and 1000M adaptive Ethernet interfaces, which can be connected to the gateway via fiber optic cable or network cable.

[0008] Preferably, the gateway device includes: The protocol conversion module parses Ethernet data packets and converts them into RS-485, CAN, and Modbus protocols, and then converts them in reverse to transmit data back from the well. The data aggregation module collects sensor data from the industrial ring network in real time and encapsulates it into Ethernet backhaul frames; The voice relay module identifies and forwards voice data to the target extension unit; Dual-port industrial switch, supporting ring topology redundancy and link fault self-healing.

[0009] Preferably, the gateway device further includes a dynamic bandwidth allocation method based on voice activity detection and load prediction, comprising the following steps: Voice activity detection: Real-time analysis of voice data streams from ground devices, calculation of short-time energy and zero-crossing rate of voice frames, comparison with preset energy thresholds and zero-crossing rate thresholds to generate voice activity flags to distinguish between valid voice and silent periods; Load forecasting: Historical load data of the industrial ring network is collected, and the predicted load in the near future is calculated using the exponential smoothing method, where the smoothing factor is dynamically adjusted based on the stability of the industrial ring network. Dynamic time slot allocation: Based on the voice activity probability calculated from the voice activity flag and the normalized value of the predicted load, combined with the total number of time-division multiplexed time slots and the number of fixed time slots reserved for control commands, the number of time slots allocated to voice data is dynamically determined through nonlinear calculation. In the dynamic time slot allocation step, the number of time slots S allocated to speech is... alloc Calculated using the following formula: ; Among them, S voice S represents the remaining time slots. voice =S total -S ctrl S total S represents the total number of time slots reused during time division; ctrl Indicates the number of control command time slots; P vad S represents the probability of voice activity; alloc Indicates the actual number of allocated speech time slots; L pred Indicates the predicted load; Time slot adjustment feedback: The actual end-to-end delay of the voice data is measured, compared with the target delay to obtain an adjustment factor, and this factor is used to perform feedback correction on the allocated number of voice time slots, forming a closed-loop control. ; ; in, Indicates the adjusted number of speech time slots; D actual D represents the actual end-to-end delay of voice data from the ground station to the extension unit; target β represents the target delay; β represents the adjustment factor.

[0010] Preferably, the extension unit includes: The bus level conversion unit provides RS-485, CAN, and Ethernet physical interfaces to realize bidirectional conversion between bus signals and logic levels; The voice processing link is equipped with a 300Hz-3.4kHz hardware filter to suppress ambient noise and integrates a ≥10W Class D power amplifier to drive the explosion-proof speaker. The input / output unit acquires digital switch signals and issues analog control commands. The wireless receiver unit has a built-in 2.4GHz WiFi and ISM band multi-mode radio frequency module, which can extend the mobile terminal access and roaming function to 200 meters.

[0011] Preferably, the roaming handover performed by the wireless receiving unit further includes a seamless roaming method based on game theory power adjustment and channel prediction, the method comprising: B1. Distributed Power Adjustment Game: Each sub-unit acts as a participant in the game, independently adjusting its wireless transmission power within power constraints to maximize its own utility function U. i (P i ,P -i ): ; Where N represents the total number of active extension units in the system; P -i P represents a vector containing the power of all N-1 units except unit i; iP represents the transmit power of extension unit i. i ∈[P min ,P max ]; P min and P max These represent the minimum and maximum values ​​of the transmitter power of the sub-unit, respectively; g ii Indicates the channel gain from unit i to its currently serving mobile terminal; g ij Nj represents the interference channel gain generated by unit j on the terminal served by unit i; N0 represents the ambient noise power spectral density. and Represents the weighting coefficients, satisfying ; B2. Channel Quality Prediction: Obtain the historical received signal strength indication sequence of the mobile terminal collected by the gateway device, and calculate the probability of the terminal accessing each candidate extension unit in the next moment based on the Hidden Markov Model. B3. Predictive-assisted seamless handover: When the current received signal strength of the mobile terminal is lower than the handover threshold and the predicted probability of it accessing a candidate extension unit in the next moment is higher than the probability threshold, the gateway instructs the candidate extension unit to establish a communication link in advance to complete the handover.

[0012] Preferably, the industrial bus protocol includes RS-485, CAN, and Modbus protocols; The gateway device adapts to multi-protocol terminal interfaces through a bus level conversion unit.

[0013] Preferably, the sub-unit collects data from the connected sensors in real time, transmits it to the gateway device through the industrial ring network, and the gateway encapsulates it into Ethernet data packets and sends them back to the ground.

[0014] Preferably, the ring topology redundancy is as follows: when a break occurs in the industrial ring network, the dual-port industrial switch switches to the backup path within ≤50ms.

[0015] Preferably, the processor unit of the ground-end device collaboratively processes sensor data and industrial control logic operations, and the audio codec module of the gateway device performs low-loss codec processing on the voice signal.

[0016] Compared with the prior art, the above-mentioned technical solution of the present invention has the following beneficial technical effects: This invention designs a real-time digital voice intercom system based on an industrial fieldbus. By deeply integrating the industrial fieldbus with real-time digital voice communication, it achieves synchronous and reliable transmission of control commands and high-fidelity voice, significantly improving the real-time performance and coordination of communication in industrial scenarios. It employs a ring topology redundancy and a rapid self-healing mechanism to ensure high network reliability during link failures, effectively adapting to harsh industrial environments such as high noise and explosion-proof environments. Through a dynamic bandwidth allocation method based on voice activity detection and load prediction, the system intelligently optimizes bus resource utilization, avoiding voice interruptions and ensuring low-latency transmission. The extension unit integrates hardware noise suppression and explosion-proof design, enhancing voice clarity and safety. Simultaneously, the system supports multi-protocol compatibility and seamless roaming functionality based on game theory and channel prediction, enabling flexible access and smooth switching of mobile terminals and expanding the operational coverage. The overall system has the advantages of high integration, flexible deployment, and controllable cost, meeting the urgent needs of modern industrial automation for efficient and reliable communication. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the technical architecture of a real-time digital voice intercom system based on an industrial fieldbus proposed in this invention. Figure 2 This is a flowchart illustrating the working mode of a real-time digital voice intercom system based on an industrial fieldbus, as proposed in this invention. Detailed Implementation

[0018] Example 1: The present invention proposes a real-time digital voice intercom system based on an industrial fieldbus, the technical principle of which is as follows: Figure 1 As shown, it includes: ground terminal equipment, gateway equipment, and extension units.

[0019] The ground-based equipment, acting as the voice control hub, collects voice through a microphone array and converts it into digital signals. After being encoded by a processor, the signals are encapsulated into Ethernet frames containing control commands and transmitted to the gateway via an adaptive network port. At the same time, it receives and restores downhole audio signals to drive the speaker output. The ground-based equipment includes audio input / output units, a processor unit, and a network port PHY module, specifically: The audio input / output unit acquires voice signals through a noise-canceling microphone array, converts them into digital signals via an ADC, and simultaneously restores the received digital voice to analog audio via a DAC to drive the speaker output, with built-in noise suppression functionality. The processor unit integrates a digital signal processor (DSP) to encode and decode voice signals, encapsulate data frames (including 8-byte control instruction fields), and collaboratively process sensor data and industrial control logic operations. The network port PHY module provides a 10 / 100 / 1000M adaptive Ethernet physical layer interface, which can be connected to the gateway device through fiber optic or network cable to realize high-speed data interaction and protocol adaptation between the ground end and the industrial ring network; The gateway device enables bidirectional conversion between industrial bus and Ethernet protocols, parses ground-side data packets and distributes them (voice relay to extension units / control commands are sent to execution devices), and simultaneously aggregates sensor data feedback. It also features a built-in dual-port redundant switch to ensure self-healing of ring network faults. Gateway devices include a protocol conversion module, a data aggregation module, a voice relay module, a hardware function module, and an audio codec module; The protocol conversion module parses the Ethernet data packets sent from the ground end, converts them into industrial bus protocols (RS-485 / CAN / Modbus), and reverses them to convert the data transmitted back from the downhole equipment. The data aggregation module collects sensor data (such as gas concentration and temperature) in the industrial ring network in real time, encapsulates it into Ethernet data packets, and transmits it back to the ground. The voice relay module identifies and forwards voice data to the target extension unit, and synchronously drives the downhole audio output device to achieve voice broadcasting. Hardware functional modules: Dual-port industrial switches support ring topology redundancy and link self-healing (fault switching ≤50ms); bus level conversion unit provides multi-protocol terminal interfaces. The audio codec module works in conjunction with the sub-unit to perform low-loss encoding and decoding of the voice signal, ensuring high-fidelity transmission. The extension unit receives voice data forwarded by the gateway and drives the explosion-proof speaker output. It collects equipment switch signals and sensor signals, interacts with industrial equipment through a bus level conversion interface, and has a built-in multi-mode wireless module to extend the access range of mobile terminals and relay functions to 200 meters. The extension unit includes a bus level conversion unit, a voice processing link, an input / output unit, and a wireless receiving unit, specifically: The bus level conversion unit provides RS-485 / CAN / Ethernet physical interfaces, is compatible with industrial equipment protocols, and realizes bidirectional conversion between bus signals and internal logic levels; The voice processing link includes an audio input unit that uses a 300Hz-3.4kHz hardware filter to suppress ambient noise; and an audio output unit that integrates a ≥10W Class D amplifier to drive an explosion-proof speaker. Input / output unit: digital input acquires device switch and alarm signals; analog output sends control commands such as valve opening and motor speed. The wireless receiving unit has a built-in 2.4GHz WiFi and ISM band multi-mode radio frequency module, which can be used as a base station relay to extend the coverage of mobile terminals to 200 meters and supports roaming switching.

[0020] Example 2: This invention proposes a real-time digital voice intercom system based on an industrial fieldbus, such as... Figure 2 As shown, its working method specifically includes the following steps: S1. The ground-end equipment performs voice acquisition and initial processing, specifically: the microphone array of the ground-end equipment acquires analog voice signals, suppresses environmental noise through a pre-bandpass filter (300Hz-3.4kHz), and then converts them into digital voice signals through an analog-to-digital converter (ADC); the processor unit encodes the digital voice and encapsulates the encoded voice data and control commands (with a reserved 8-byte field) into Ethernet data frames, which are then sent to the gateway device through the network port PHY module.

[0021] S2. The gateway device performs data transmission and protocol parsing. Specifically, after receiving Ethernet data frames, the gateway device separates voice data and control commands: the voice data is directly forwarded to the audio codec module; the control commands are parsed into the target industrial bus protocol (RS-485 / CAN / Modbus) by the protocol conversion module and converted into physical layer signals by the bus level conversion unit; at the same time, the gateway continuously collects sensor data (such as gas concentration and temperature) in the industrial ring network and encapsulates it into Ethernet return frames.

[0022] S3. The gateway device and the extension unit collaboratively perform dynamic allocation of bus resources. The gateway device further adopts a dynamic bus resource allocation method based on voice perception and load entropy state prediction. By monitoring voice activity in real time, predicting network load, and combining feedback adjustments, it dynamically optimizes the allocation of time-division multiplexing time slots. The specific implementation process is as follows: A1. The gateway device monitors the voice data stream from the ground terminal device in real time and uses a Voice Activity Detection (VAD) algorithm to determine whether there is a valid voice signal. This is achieved by calculating the short-time energy and zero-crossing rate of the voice frame to distinguish between voice and background noise, thereby avoiding wasting bandwidth resources during silent periods. ; Where V(n) is a binary flag, 1 indicates speech activity, and 0 indicates silence; Z(n) is the zero-crossing rate of the nth frame, i.e., the number of times the signal crosses zero points within the frame, reflecting the signal frequency characteristics; E(n) is the short-time energy of the nth frame, calculated by the sum of squares of the sampled values ​​within the frame, reflecting the signal strength; E th Z represents the energy threshold, set based on background noise statistics to suppress sudden noise; th This indicates the typical speech frequency range (300Hz-3.4kHz); values ​​that are too high usually indicate noise. A2. The gateway device continuously collects historical load data of the industrial ring network (such as the number of data packets per unit time) and predicts the network load in the near future based on the exponential smoothing method. This allows the system to adjust resource allocation in advance to cope with load fluctuations and avoid congestion. The prediction model is as follows: ; Where L(t) represents the actual load at time t, which is statistically analyzed in real time by the gateway data aggregation module, reflecting the current network traffic; L pred (t) represents the predicted load, used to estimate future load trends. Its value is updated recursively through exponential smoothing to balance the latest data and historical trends. α represents the smoothing factor, which controls the sensitivity of the prediction to recent changes. In this embodiment, the value is 0.1-0.3, and the stability of the industrial ring network is determined by simulation. A3. Based on VAD output and load prediction, the gateway device dynamically calculates the number of time slots allocated to voice data in each time-division multiplexing cycle, prioritizes the fixed time slots of control commands (high priority), and then adjusts the voice time slots according to the probability of voice activity and predicted load to ensure that the end-to-end latency is ≤50ms. Number of time slots S allocated to speech alloc for: ; Among them, S voice S represents the remaining time slots. voice =S total -S ctrl S total The total number of time-division multiplexed time slots is determined by the bus bandwidth and cycle length; S ctrl This indicates the number of control command time slots, set according to the maximum number of control devices and safety redundancy; P vad S represents the probability of speech activity, reflecting recent speech activity levels; alloc This indicates the actual number of voice time slots allocated. Over-allocation is avoided through a minimum function, ensuring that the bus is not overloaded. A4. The gateway device introduces a feedback mechanism to adjust time slot allocation based on the measured delay. By comparing the actual delay with the target delay, the allocation strategy is dynamically corrected to form a closed-loop control. Define the actual end-to-end delay D of voice data from the ground station to the extension unit. actual Target delay D target (In this embodiment, the value is 50ms), and the adjustment factor β is: ; Then update the voice time slot allocation: ; in, This indicates the adjusted number of voice time slots, ensuring that the allocation does not exceed the remaining time slot capacity.

[0023] S4. The sub-unit performs underground voice broadcasting and control execution. Specifically, the sub-unit receives voice data sent by the gateway, converts it back into analog audio signals via a digital-to-analog converter (DAC), and drives the explosion-proof speaker output by a Class D power amplifier (output power ≥10W). At the same time, the sub-unit collects equipment status signals (such as switch signals and alarm signals) through the digital input interface, and sends control commands (such as adjusting valve opening and motor speed) through the analog output interface to execute the control commands forwarded by the gateway.

[0024] S5. The sub-unit and the gateway work together to transmit sensor data back. Specifically, the sub-unit collects the connected sensor data in real time and transmits it to the gateway through the industrial ring network. The gateway encapsulates the sensor data into Ethernet data packets and transmits them back to the ground equipment via the network port PHY module for processing by the monitoring system.

[0025] S6. The wireless module of the extension unit performs wireless relay access and roaming. Specifically, the wireless receiving unit of the extension unit (supporting 2.4GHz WiFi / ISM bands) further executes a seamless roaming method based on game theory power adjustment and channel prediction. It suppresses co-channel interference through distributed power game theory, predicts the trajectory of the mobile terminal using a hidden Markov model, and triggers prediction-assisted seamless handover. Specifically: B1. Each sub-unit, as a participant in the game, independently adjusts its wireless transmission power to seek the optimal balance between coverage and interference. The goal of each unit is to maximize its own utility function, that is: the utility function U of any hierarchical unit i. i (P i ,P -i )for: ; Where N represents the total number of active extension units in the system, maintained and broadcast by the gateway device; P -i P represents a vector containing the power of all N-1 units except unit i; i P represents the transmit power of extension unit i (where i∈[1,N]). i ∈[P min ,P max ]; P min and P max These represent the minimum and maximum values ​​of the transmitter power of the sub-unit, respectively; g ii Indicates the channel gain from unit i to its currently serving mobile terminal; g ij The interference channel gain generated by unit j on the terminal served by unit i is obtained through the periodic exchange of channel state information (CSI) between extension units; N0 represents the ambient noise power spectral density, which is measured by the extension unit in the idle time slot. and These represent weighting coefficients, respectively characterizing the importance of signal quality (related to communication capacity) and power consumption. They are set offline based on network deployment density and energy consumption requirements to meet [the following criteria]. ; B2. Obtain the historical RSSI (Received Signal Strength Indication) sequences of each mobile terminal collected by the gateway device, train a Hidden Markov Model (HMM), and predict the best extension unit and channel quality that the terminal may access in the future, i.e.: Define observation sequence O 1:t ={O1,O2,…,O t Let} be the historical RSSI values ​​over the past t time points, and let the hidden state sequence be S={S1,S2,…,S…} t} represents the extension unit ID associated with the terminal in the past t time intervals, predicting the probability that the terminal will be in state k (i.e., access hierarchical unit k) in the next time interval t+1. for: ; Where t is the current time point; O 1:t This represents the observation sequence from time 1 to the current time t; M represents the total number of hidden states, i.e., the number of extension units that the mobile terminal may access, M≤N; r,k are state indices, r,k∈[1,M]; The state transition probability represents the probability of moving from the coverage area of ​​extension unit r to the coverage area of ​​extension unit k, and is learned from historical movement data through the Baum-Welch algorithm (forward-backward algorithm). Given the observation sequence up to time t, this represents the probability (posterior probability) of being in state r at the current time, which is recursively calculated using the forward algorithm; B3. Combining the power game results and channel prediction, when the mobile terminal's RSSI is below the threshold and the predicted probability is above the threshold, the gateway instructs the target extension unit to allocate resources and adjust power in advance, achieving seamless handover with "connect first, then disconnect". The handover decision condition is: ; Among them, RSSI current This indicates the received signal strength measured by the current serving extension unit of the mobile terminal; T handover This represents the traditional handover threshold, set based on experience in industrial environments; K represents the set of candidate extension units, which in this embodiment is determined by the gateway after excluding units with weak signals or excessive loads based on the network topology. ; This indicates that, in the candidate set K, the value with the highest probability of access at the next time step is selected; P th This represents the predicted probability threshold, which is adjusted through simulation to balance switching timeliness and misjudgment rate. Accordingly: the passive response handover is transformed into an active predictive handover, by selecting the best target from a clear set of candidate units and establishing the target link in advance, avoiding voice interruption and ensuring roaming latency ≤100ms.

[0026] S7. Industrial ring network performs fault recovery and network redundancy. Specifically, when an industrial ring network (fiber optic / shielded twisted pair) experiences a break, the ring topology initiates a self-healing mechanism: the dual-port industrial switch switches to the backup path within ≤50ms to ensure the continuous transmission of voice and control signals.

[0027] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited thereto. Various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention.

Claims

1. A real-time digital voice intercom system based on an industrial fieldbus, characterized in that, include: Ground-based equipment is used to collect voice signals through a microphone array, encode them, encapsulate them with control commands into Ethernet frames, and send them to the gateway device. It also receives and restores the audio signals to drive the speaker output. Gateway devices are used to realize bidirectional conversion between industrial bus protocols and Ethernet, separate voice data and control commands, forward voice data to the target extension unit, convert control commands into industrial bus protocols and send them to the execution device, collect sensor data and send it back, and dynamically allocate bus resources using a time-division multiplexing mechanism. The gateway device further executes a dynamic bandwidth allocation method based on voice activity detection and load prediction, including the following steps: Voice activity detection: Real-time analysis of voice data streams from ground devices, calculation of short-time energy and zero-crossing rate of voice frames, comparison with preset energy thresholds and zero-crossing rate thresholds to generate voice activity flags to distinguish between valid voice and silent periods; Load forecasting: Historical load data of the industrial ring network is collected, and the predicted load in the near future is calculated using the exponential smoothing method, where the smoothing factor is dynamically adjusted based on the stability of the industrial ring network. Dynamic time slot allocation: Based on the voice activity probability calculated from the voice activity flag and the normalized value of the predicted load, combined with the total number of time-division multiplexed time slots and the number of fixed time slots reserved for control commands, the number of time slots allocated to voice data is dynamically determined through nonlinear calculation. In the dynamic time slot allocation step, the number of time slots S allocated to speech is... alloc Calculated using the following formula: ; Among them, S voice S represents the remaining time slots. voice =S total -S ctrl S total S represents the total number of time slots reused during time division; ctrl Indicates the number of control command time slots; P vad S represents the probability of voice activity; alloc Indicates the actual number of allocated speech time slots; L pred Indicates the predicted load; Time slot adjustment feedback: The actual end-to-end delay of the voice data is measured and compared with the target delay to obtain an adjustment factor. This adjustment factor is then used to perform feedback correction on the allocated number of voice time slots, forming a closed-loop control. ; ; in, Indicates the adjusted number of speech time slots; D actual This represents the actual end-to-end delay of voice data from the ground station to the extension unit; D target β represents the target delay; β represents the adjustment factor. The extension unit is used to receive voice data to drive the output of the explosion-proof speaker, collect equipment switching signals and sensor signals, and supports wireless relay access and roaming switching.

2. The real-time digital voice intercom system based on industrial fieldbus according to claim 1, characterized in that, Ground-based equipment includes: Audio input / output unit, equipped with a noise-canceling microphone array and analog-to-digital converter and digital-to-analog converter; The processor unit integrates a DSP chip to perform encoding and decoding, and encapsulates data frames containing 8-byte control instruction fields; The network port PHY module provides 10, 100, and 1000M adaptive Ethernet interfaces, which can be connected to the gateway device via fiber optic cable or network cable.

3. The real-time digital voice intercom system based on industrial fieldbus according to claim 1, characterized in that, Gateway devices include: The protocol conversion module parses Ethernet data packets and converts them into RS-485, CAN, and Modbus protocols, and then converts them in reverse to transmit data back from the well. The data aggregation module collects sensor data from the industrial ring network in real time and encapsulates it into Ethernet backhaul frames; The voice relay module identifies and forwards voice data to the target extension unit; Dual-port industrial switch, supporting ring topology redundancy and link fault self-healing.

4. A real-time digital voice intercom system based on an industrial fieldbus according to claim 1, characterized in that, The extension unit includes: The bus level conversion unit provides RS-485, CAN, and Ethernet physical interfaces to realize bidirectional conversion between bus signals and logic levels; The voice processing link is equipped with a 300Hz-3.4kHz hardware filter to suppress ambient noise and integrates a ≥10W Class D power amplifier to drive the explosion-proof speaker. The input / output unit acquires digital switch signals and issues analog control commands. The wireless receiver unit has a built-in 2.4GHz WiFi and ISM band multi-mode radio frequency module, which can extend the mobile terminal access and roaming function to 200 meters.

5. A real-time digital voice intercom system based on an industrial fieldbus according to claim 4, characterized in that, The roaming handover performed by the wireless receiver unit further includes a seamless roaming method based on game theory power adjustment and channel prediction, specifically including: B1. Distributed Power Adjustment Game: Each sub-unit acts as a participant in the game, independently adjusting its wireless transmission power within power constraints to maximize its own utility function U. i (P i ,P -i ): ； Where N represents the total number of active extension units in the system; P -i P represents a vector containing the power of all N-1 extension units except extension unit i; i P represents the transmit power of extension unit i. i ∈[P min ,P max ]; P min and P max These represent the minimum and maximum values ​​of the transmitter power of the sub-unit, respectively; g ii Indicates the channel gain from extension unit i to its currently serving mobile terminal; g ij Nj represents the interference channel gain generated by extension unit j on the mobile terminal served by extension unit i; N0 represents the ambient noise power spectral density. and Represents the weighting coefficients, satisfying ; B2. Channel Quality Prediction: Obtain the historical received signal strength indication sequence of the mobile terminal collected by the gateway device, and calculate the probability of the mobile terminal accessing each candidate extension unit in the next moment based on the Hidden Markov Model. B3. Predictive-assisted seamless handover: When the current received signal strength of the mobile terminal is lower than the handover threshold, and the predicted probability of it accessing a candidate extension unit in the next moment is higher than the probability threshold, the gateway device instructs the candidate extension unit to establish a communication link in advance to complete the handover.

6. A real-time digital voice intercom system based on an industrial fieldbus according to claim 1, characterized in that, Industrial bus protocols include RS-485, CAN, and Modbus protocols; The gateway device adapts to multi-protocol terminal interfaces through a bus level conversion unit.

7. A real-time digital voice intercom system based on an industrial fieldbus according to claim 1, characterized in that, The sub-unit collects data from the connected sensors in real time and transmits it to the gateway device via the industrial ring network. The gateway device then encapsulates the data into Ethernet packets and sends them back to the ground terminal device.

8. A real-time digital voice intercom system based on an industrial fieldbus according to claim 3, characterized in that, Ring topology redundancy means that when a break occurs in the industrial ring network, the dual-port industrial switch will switch to the backup path within ≤50ms.

9. A real-time digital voice intercom system based on an industrial fieldbus according to claim 1, characterized in that, The processor unit of the ground-side equipment collaboratively processes sensor data and industrial control logic operations, while the audio codec module of the gateway equipment performs low-loss codec processing on the voice signal.

Images