Method for simulating converged communication of hostless sound amplification talkback system and IP network sound amplification talkback system

Abstract

The invention relates to the technical field of sound amplification talkback communication, in particular to a method for simulating converged communication of a hostless sound amplification talkback system and an IP (Internet Protocol) network sound amplification talkback system, which realizes bidirectional conversion and protocol adaptation of an analog signal and an IP digital signal by constructing a three-layer converged architecture and utilizing a converged communication gateway. According to the method, the problem of interface difference is solved by adopting impedance conversion and a two-wire and four-wire conversion technology, automatic recognition of voice instructions is realized through a voice recognition technology, and the communication quality is ensured by adopting a specific signal processing algorithm and a protocol conversion technology. The system effectively solves the problem of communication barriers between a traditional simulation system and an IP network system, has the advantages of being low in cost, good in compatibility, high in reliability and the like, and is particularly suitable for the safety production communication requirements in the industrial fields of petrochemical engineering and the like.

Description

Technical Field

[0001] This invention relates to the field of public address communication technology, specifically a method for realizing the integrated communication between an analog hostless public address system and an IP network public address system. Background Technology

[0002] Currently, in the field of industrial communication, traditional analog intercom systems are widely used in petroleum, chemical, and other scenarios due to their characteristics such as no central node, strong anti-interference capabilities, and suitability for harsh industrial environments. However, while the traditional analog signal transmission method relying on cables is simple and direct, it has significant limitations such as limited functionality and difficulty in integrating with modern digital network systems. With the deep penetration of IP network technology into various fields, the digital upgrade of industrial communication has become an inevitable trend. Against this backdrop, converged communication gateways have emerged. Their core function is to receive analog signals and convert them into digital signals. Through a built-in network communication unit, they establish an Ethernet connection with the IP network intercom system server, realizing digital transmission and intelligent processing of signals. At the same time, the gateway has bidirectional conversion capabilities, which can restore digital signals to analog signals, allowing traditional analog equipment to seamlessly connect to the digital network environment without modification, fundamentally breaking down the communication barriers between analog and digital systems.

[0003] Traditional oil and petrochemical companies often have analog hostless intercom systems. However, during the current digital transformation and upgrading process, these companies simultaneously have both analog hostless intercom systems and IP network intercom systems. The two systems are protocol-incompatible and cannot communicate directly. Replacing the old analog hostless intercom system requires replacing all equipment, resulting in high costs. Therefore, developing a converged communication gateway has become essential. Summary of the Invention

[0004] The purpose of this invention is to provide a method for realizing the integrated communication between an analog hostless public address intercom system and an IP network public address intercom system, so as to solve the problem mentioned in the background art that existing analog systems and modern IP network systems cannot communicate directly due to differences in protocols and interfaces.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A method for achieving integrated communication between an analog hostless public address intercom system and an IP network public address intercom system, wherein the method realizes bidirectional conversion and protocol adaptation of analog signals and IP digital signals through a unified communication gateway, including the following steps: (1) Construct a three-layer converged architecture consisting of an IP network-based public address intercom system, a converged communication gateway, and an analog public address intercom system; (2) The converged communication gateway receives two-wire voice signals from analog telephone stations, converts the two-wire signals into four-wire signals through impedance transformation and two-to-four-wire conversion modules, and performs impedance matching processing. (3) The voice signal is recognized and processed by the voice recognition control module, and the voice command is converted into a text command by an automatic voice recognition algorithm; (4) The core control module parses text instructions and controls the VOIP analog-to-digital signal conversion module to convert analog signals into digital VOIP signals; (5) Encapsulate the digital signal into IP data packets and transmit them to the IP network type public address system server via Ethernet; (6) When implementing reverse communication, the digital signal from the IP network is converted into an analog signal and transmitted to the analog telephone station system; The automatic speech recognition algorithm used in this process is based on a Hidden Markov Model, and its acoustic model probability calculation formula is as follows:

[0006] Where O is the observation sequence, λ is the model parameters, and q is the state sequence.

[0007] Preferably, the impedance matching process of the impedance transformation and two-to-four-wire conversion module adopts the impedance transformation formula: Z in =Z out ×(N1 / N2) 2 ; Among them, Z in Z is the input impedance. out N1 represents the output impedance, and N2 represents the turns ratio of the transformer.

[0008] Preferably, the automatic speech recognition algorithm used by the speech recognition control module also includes a language model based on a deep neural network, and its word prediction probability formula is: ; Where, ω i For the current word, ω1, ω2, ..., ω i-1 For historical vocabulary sequence, h i Here, W represents the hidden layer state, and b represents the model parameters.

[0009] Preferably, the VOIP analog-to-digital signal conversion module uses the G.711 encoding protocol for analog-to-digital conversion, and its quantization formula is as follows: ; Where x is the amplitude of the input analog signal, y is the output digital signal, and μ is the compression parameter, which is usually taken as 255.

[0010] Preferably, the core control module dynamically controls the signal paths of each module through level signals. Its control logic adopts a finite state machine model, and the state transition function is expressed as follows: δ:S×Σ→S; Where S is the set of states, Σ is the input alphabet, and δ is the state transition function.

[0011] Preferably, the method further includes an emergency linkage processing step: when the fire alarm gateway triggers an emergency signal, the converged communication gateway synchronously captures the IP-side signaling, converts the dispatch command into an analog signal, and triggers the analog-side telephone station and speaker to respond.

[0012] Preferably, the emergency response mechanism employs a priority scheduling algorithm, with the priority allocation function being: Priority = α × T emergency +β×S importance ; Among them, T emergency S is the urgency parameter. importance Here, α and β are the signal importance parameters, and α and β are the weighting coefficients.

[0013] Preferably, the VoIP analog-to-digital signal conversion module supports the SIP protocol stack, its signaling encapsulation format conforms to the RFC3261 standard, and the session establishment time meets the following requirements: T setup ≤150ms; Among them, T setup This represents the total latency from signaling to session establishment.

[0014] Preferably, the speech recognition control module achieves the following recognition accuracy in noisy environments: Accuracy ≥ 95% @ SNR ≥ 15dB; Where SNR is the signal-to-noise ratio and Accuracy is the voice command recognition accuracy.

[0015] Preferably, the converged communication gateway supports multi-channel concurrent processing, and its maximum number of concurrent channels N is determined by the following formula: ; Among them, T processing For processor processing power, T frame For the processing time per frame, C overhead This represents the system overhead coefficient.

[0016] Compared with the prior art, the beneficial effects of the present invention are: The most direct and beneficial effect of this invention lies in its outstanding economic efficiency. This method, through a dedicated converged communication gateway, perfectly realizes the reuse and transformation of existing analog systems, eliminating the need for enterprises to discard and replace existing, heavily invested facilities such as analog telephone stations, speakers, and cables scattered throughout production areas. This directly avoids the enormous hardware procurement and engineering installation costs associated with complete equipment replacement in traditional upgrade solutions, significantly reducing the barriers to entry and initial investment for the digital transformation and upgrading of communications in large enterprises such as those in the petroleum and chemical industries, making it a high-performance, cost-effective solution.

[0017] At the functional level, this invention completely breaks down the protocol barriers between analog and digital systems, bringing about a qualitative leap in management efficiency. It enables traditional, closed analog systems to seamlessly connect to modern IP networks, empowering them to achieve unified scheduling, broadcasting, and plant-wide emergency response. When the central control room issues commands via the IP network or the fire alarm system triggers, the commands can be delivered indiscriminately to every analog or digital terminal, greatly improving the command efficiency and emergency response speed for safe production, and solving the "information silo" problem caused by the coexistence of multiple systems in an enterprise.

[0018] The advanced technology of this invention forms a solid foundation for the aforementioned effects. Its converged communication gateway integrates core modules such as analog-to-digital signal conversion, intelligent speech recognition (ASR), and protocol parsing and adaptation. It not only solves the technical challenges of signal and protocol compatibility but also overcomes the operational limitations of traditional physical buttons through voice interaction, making communication operations more intuitive and convenient. This technical architecture ensures compatibility with existing systems while possessing good scalability, leaving room for future technological evolution. Ultimately, it meets the stringent requirements of industrial scenarios for communication systems in a highly reliable and low-risk manner. By achieving bidirectional conversion and protocol adaptation between analog and IP digital signals through the converged communication gateway, it breaks down communication barriers between two systems in a low-cost and low-risk way, meeting the core needs of cross-regional dispatching and emergency response in enterprise safe production. Attached Figure Description

[0019] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are explained in detail together with the embodiments of the invention, but do not constitute a limitation thereof.

[0020] Figure 1 This invention provides a block diagram of the integrated communication system that combines a hostless analog intercom system with an IP network intercom system. Figure 2 This is a bar chart comparing the required values ​​and measured values ​​of key performance indicators in Embodiment 1 of the present invention. Figure 3 This is a line graph showing the changing trend of speech recognition rate under different signal-to-noise ratio (SNR) environments in Embodiment 2 of the present invention. Figure 4 This is a scatter plot of the end-to-end response time for each of the ten repeated tests in Embodiment 3 of the present invention. Detailed Implementation

[0021] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0022] This invention achieves integrated communication between an analog hostless public address system and an IP network public address system through an analog-to-digital signal conversion device, as detailed below: This invention constructs a three-layer converged architecture consisting of an "IP network-based public address intercom system → converged communication gateway → analog public address intercom system". By using the converged communication gateway, it breaks through the protocol and signal format barriers between the two types of systems, supporting core scenarios such as safe production scheduling and emergency response in enterprises such as petroleum and chemical industries.

[0023] IP network-based public address intercom systems mainly consist of digital telephone stations, loudspeakers, fire alarm gateways, and system servers. They are networked via LAN switches, transmit digital signals over IP networks, support standardized protocols such as SIP, and have integrated management capabilities such as digital dispatching and fire linkage. Analog public address intercom systems use loudspeakers and analog telephone stations networked via bus cables. They are easy to deploy, lightweight, and have low maintenance costs, but their closed protocols make them difficult to interface with IP network systems. The two types of systems interact through a converged communication gateway.

[0024] The converged communication gateway integrates a VoIP analog-to-digital signal conversion module, an impedance transformation and two-to-four-wire conversion module, a core control module, and a voice recognition module. Among them, the VoIP analog-to-digital signal conversion module is responsible for bidirectional conversion between analog signals and digital VoIP signals, realizing the encapsulation and decapsulation of digital signals, and parsing analog proprietary protocols and IP standard protocols, solving the pain point of "protocol incompatibility preventing direct connection". The analog telephone station uses a two-wire system, with the same pair of cables simultaneously enabling both "receiving" and "transmitting". The gateway's internal processing is more adapted to a four-wire system. The impedance transformation and two-to-four-wire conversion module enables bidirectional conversion between two-wire and four-wire systems, while also matching the impedance characteristics of the analog telephone station and the VoIP module to solve the interface difference problem.

[0025] The core control module undertakes logic scheduling and protocol interaction functions, and uses level signals to dynamically control the signal paths of the VoIP analog-to-digital signal conversion module, impedance transformation and two-to-four-wire conversion module, and voice recognition module.

[0026] The voice recognition control module receives the pre-processed voice signal and converts the voice into text commands, such as "call digital station 1", through the ASR (Automatic Speech Recognition) algorithm. The recognition result is then transmitted to the core control module, breaking through the limitations of traditional intercom "button operation" and realizing voice interaction.

[0027] Operation procedure for analog public address system calling IP network public address system: When analog base station 1 goes off, the user inputs a voice command. The analog audio signal is transmitted to the converged communication gateway via a two-wire bus cable. At this time, the voice signal acts as a trigger source, first activating the gateway's voice recognition control module. After the module starts, it sends a "voice command detected" signal to the core control module through its internal wake-up line. The core control module responds and starts, outputting level control signals through interfaces K1 and K3 to trigger the impedance transformation and two-to-four-wire conversion module and the VoIP analog-to-digital signal conversion module to switch from standby to operating state. Simultaneously, it sends a shutdown command to the voice recognition control module through K5.

[0028] After the impedance transformation and 2-to-4-wire conversion module is activated, it converts the input 2-wire voice signal into a 4-wire signal. The impedance is then adjusted by an impedance matching circuit to ensure the analog station's output impedance matches the gateway's internal circuit impedance. The processed 4-wire voice signal is transmitted via a dedicated line to the voice recognition control module, which has already received the shutdown command. Before the shutdown command takes effect, the voice recognition control module uses the ASR algorithm to convert the voice signal into a text command, "Calling digital station 1," which is transmitted to the core control module via the K4 interface. After parsing the command, the core control module outputs a level signal through the K1 interface to activate the VoIP analog-to-digital converter module. The VoIP module encapsulates the signaling into IP data packets, which are then forwarded to the system server of the IP network-type public address system via a LAN switch. The system server parses the IP data packets, locates digital station 1, and initiates a call. Digital station 1 answers the call to complete the conversation.

[0029] Operation procedure for calling an analog public address system from an IP network public address system: Digital station 1 generates an activation signal via dialing commands, which is then forwarded by the switch to the VoIP analog-to-digital converter module of the converged communication gateway, triggering gateway startup. The user then inputs a voice command to use a specific channel; this digital signal is transmitted via the server and switch to the voice recognition control module. The voice recognition control module parses the command using the ASR algorithm and transmits it to the core control module via the K4 interface. The core control module outputs a level signal through the K1 interface to switch internal paths. The user broadcasts through the microphone of digital station 1, and the analog user picks up the phone, switching to the responding channel, thus completing the intercom.

[0030] The core value of this solution lies in: Cost dimension: By "reusing and upgrading existing equipment", the complete replacement of analog base stations is avoided, significantly reducing hardware investment for digital transformation; Functional dimension: Break down communication barriers between the two systems, support cross-regional dispatch, fire-fighting linkage and other scenarios, and improve the efficiency of collaborative safety production; Technical aspects: Integrating technologies such as analog-to-digital conversion and protocol adaptation, while taking into account compatibility with existing systems and expansion into new systems, it provides low-risk, cost-effective converged communication solutions for industrial scenarios such as oil and petrochemicals.

[0031] Emergency response linkage is the core capability of this integrated solution, which is achieved by relying on the two-way interactive characteristics of the integrated communication gateway: when the fire alarm gateway triggers an emergency signal, the integrated communication gateway simultaneously captures the signaling, converts the IP-side dispatching command into an analog signal, and triggers the analog-side telephone station and speaker to respond.

[0032] 1. Example Name: Example of verifying the basic communication functions and performance of a converged communication gateway in a standard petrochemical plant environment.

[0033] 2. Objective: To verify whether the method and gateway device described in this invention can stably realize bidirectional calls between analog and IP systems in a typical industrial environment, and to test whether its key performance indicators meet the design requirements.

[0034] 3. Implementation conditions and configuration: On the analog system side: four traditional two-wire analog telephone stations are deployed, with a total bus cable length of approximately 500 meters and a line characteristic impedance of 600Ω.

[0035] On the IP system side: Two IP digital telephone stations are deployed, connected to an enterprise-grade gigabit LAN. The system server supports the SIP protocol (RFC3261).

[0036] Converged communication gateway: A prototype machine was built according to the present invention. The core processor is ARM Cortex-A53 and the voice encoding is G.711a law (μ=255).

[0037] Test environment: Central control room of a petrochemical plant, with an ambient noise level of approximately 65 dB.

[0038] 4. Implementation steps and results: Analog side calling IP side: Pick up the phone at analog station A and say the voice command "Call digital station 1". The gateway successfully recognizes the call and establishes a connection.

[0039] IP-side call to analog-side: On digital station 1, dial "Analog Area Broadcast" and say the command "Call Analog Station A". The gateway successfully triggers ringing on all analog-side stations. After station A goes off, the call is established.

[0040] Performance testing: The call process was monitored using a network tester and an audio analyzer. Key data is shown in the table below.

[0041] 5. Data Table: Performance Test Results of Example 1

[0042] 6. Results of the Example: This example fully demonstrates that the method described in this invention can effectively achieve seamless bidirectional communication between analog and IP systems. All key performance indicators have met or exceeded design expectations, satisfying the core requirements of industrial environments for real-time and reliable communication.

[0043] like Figure 2 The bar chart compares the required and measured values ​​of key performance indicators, clearly showing the achievement of each indicator. In particular, the measured results for voice latency and MOS value are better than the requirements, demonstrating the effectiveness of the solution.

[0044] 1. Example Name: Test Example of Adaptive Speech Recognition Capability for High-Noise Scenarios such as Compressor Rooms.

[0045] 2. Objective: To test the robustness of the speech recognition module in this invention under extreme industrial noise environments and to verify the effectiveness of its algorithm.

[0046] 3. Implementation conditions and configuration: Deploying an analog telephone station and a converged communication gateway near the compressor room in the factory area resulted in continuous noise levels of 85-90 dB.

[0047] To avoid interference, the IP digital station and server remain located in the central control room.

[0048] The speech recognition module enables noise suppression (NS) and automatic gain control (AGC) algorithms.

[0049] 4. Implementation steps and results: Testers, wearing dust masks, repeatedly called commands at normal, loud, and low volumes at a simulated telephone station outside the compressor room.

[0050] Record the recognition results of the voice recognition control module and compare them with preset commands.

[0051] To quantify the effect, signal-to-noise ratio (SNR) was introduced as an evaluation metric. The test results are shown in the table below.

[0052] 5. Data Table: Speech Recognition Performance Test Results of Example 2

[0053] 6. Implementation Results: This implementation demonstrates that under extremely high ambient noise conditions, the user's pronunciation significantly impacts the recognition rate. This invention, through its built-in noise suppression and gain control algorithms, effectively improves practical usability. Even with user cooperation (appropriately increasing volume), the recognition rate remains above 94%, meeting the needs of on-site emergency calls.

[0054] like Figure 3 The line graph illustrates the trend of speech recognition rate under different signal-to-noise ratio (SNR) environments. It can be seen that the recognition rate drops sharply as the SNR decreases (environmental noise worsens). However, after enabling the noise reduction (NS) and automatic gain control (AGC) algorithms, the recognition rate was significantly improved under the same adverse conditions (-5 dB SNR), increasing from approximately 75% to 94%, demonstrating the core value of the algorithm module.

[0055] 1. Example Name: Example of Fire Alarm Triggering Plant-Wide Emergency Broadcast Linkage

[0056] 2. Objective: To verify the actual effect of the present invention in the core scenario of safe production—emergency response, and to test the real-time response and reliability of the system.

[0057] 3. Implementation conditions and configuration: The analog system covers the old equipment area of ​​the plant (10 analog telephone stations / speakers).

[0058] The IP system covers the new equipment area and the central control room (5 digital telephone stations).

[0059] The factory's fire protection system is equipped with a fire alarm gateway, which can send alarm signals (standard SIPINFO messages) to the IP network intercom system server.

[0060] Emergency linkage logic is preset in the core control module of the converged communication gateway: upon receiving a specific alarm signal, a full-simulation area broadcast is immediately forced.

[0061] 4. Implementation steps and results: Simulate triggering a fire alarm (e.g., pressing a manual alarm button).

[0062] The fire alarm gateway generates an alarm signal and sends it to the IP intercom system server via the local area network.

[0063] The server immediately sends the alarm information and pre-stored voice files (such as "Fire in area 101, evacuate immediately") as dispatch instructions to the converged communication gateway.

[0064] The gateway's VoIP module receives and unpacks digital signals. The core control module prioritizes processing this instruction, controlling the audio path to convert the digital voice stream into an analog signal and outputting the maximum volume to all analog stations and speakers.

[0065] Record the end-to-end delay from alarm triggering to the start of evacuation instructions playing from the simulated zone loudspeakers. Repeat the test 10 times.

[0066] 5. Data Table: Emergency Response Performance Test in Example 3

[0067] 6. Implementation Results: This embodiment successfully verifies the key value of the present invention in emergency response scenarios. The converged communication gateway can complete the entire process from receiving digital alarm signals to driving the entire traditional analog system to conduct emergency broadcasts within an average of 805 milliseconds. It has a rapid response, high reliability, and fully meets the emergency communication needs of safe production in the petrochemical industry, solving the pain point that old systems cannot be integrated into modern intelligent alarm systems.

[0068] like Figure 4 The scatter plot shown illustrates the end-to-end response time (in milliseconds) for each of the ten repeated tests. All data points are below the 1000ms requirement and fluctuate slightly around the average of 805ms, demonstrating the high efficiency and stability of the system response, fully meeting the stringent time latency requirements of emergency scenarios.

[0069] The method of realizing the integrated communication between the analog hostless public address intercom system and the IP network public address intercom system of the present invention has the following advantages: The most direct and beneficial effect of this invention lies in its outstanding economic efficiency. This method, through a dedicated converged communication gateway, perfectly realizes the reuse and transformation of existing analog systems, eliminating the need for enterprises to discard and replace existing, heavily invested facilities such as analog telephone stations, speakers, and cables scattered throughout production areas. This directly avoids the enormous hardware procurement and engineering installation costs associated with complete equipment replacement in traditional upgrade solutions, significantly reducing the barriers to entry and initial investment for the digital transformation and upgrading of communications in large enterprises such as those in the petroleum and chemical industries, making it a high-performance, cost-effective solution.

[0070] At the functional level, this invention completely breaks down the protocol barriers between analog and digital systems, bringing about a qualitative leap in management efficiency. It enables traditional, closed analog systems to seamlessly connect to modern IP networks, empowering them to achieve unified scheduling, broadcasting, and plant-wide emergency response. When the central control room issues commands via the IP network or the fire alarm system triggers, the commands can be delivered indiscriminately to every analog or digital terminal, greatly improving the command efficiency and emergency response speed for safe production, and solving the "information silo" problem caused by the coexistence of multiple systems in an enterprise.

[0071] The advanced technology of this invention forms a solid foundation for the aforementioned effects. Its converged communication gateway integrates core modules such as analog-to-digital signal conversion, intelligent speech recognition (ASR), and protocol parsing and adaptation. It not only solves the technical challenges of signal and protocol compatibility but also overcomes the operational limitations of traditional physical buttons through voice interaction, making communication operations more intuitive and convenient. This technical architecture ensures compatibility with existing systems while possessing good scalability, leaving room for future technological evolution. Ultimately, it meets the stringent requirements of industrial scenarios for communication systems in a highly reliable and low-risk manner. By achieving bidirectional conversion and protocol adaptation between analog and IP digital signals through the converged communication gateway, it breaks down communication barriers between two systems in a low-cost and low-risk way, meeting the core needs of cross-regional dispatching and emergency response in enterprise safe production.

[0072] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for realizing fusion communication of an analog hostless loudspeaker intercom system and an IP network loudspeaker intercom system, characterized in that, The method realizes bidirectional conversion and protocol adaptation of analog signals and IP digital signals through a fusion communication gateway, and comprises the following steps: (1) constructing a three-layer fusion architecture composed of an IP network type intercom system, a fusion communication gateway and an analog intercom system; (2) the fusion communication gateway receives two-wire voice signals from an analog talk station, converts the two-wire signals into four-wire signals through an impedance conversion and two-four wire conversion module, and performs impedance matching processing; (3) the voice recognition control module performs recognition processing on the voice signals, and converts voice commands into text commands using an automatic speech recognition algorithm; (4) the core control module parses the text commands and controls the VOIP analog-digital signal conversion module to convert analog signals into digital VOIP signals; (5) the digital signals are packaged into IP data packets and transmitted to the IP network type intercom system server through Ethernet; (6) when reverse communication is realized, the digital signals from the IP network are converted into analog signals and transmitted to the analog talk station system; wherein the automatic speech recognition algorithm used by the voice recognition module is based on a hidden Markov model, and the acoustic model probability calculation formula is:

2. wherein O is an observation sequence, λ is a model parameter, and q is a state sequence.

3. The method of claim 1, wherein the method further comprises: The impedance matching processing of the impedance conversion and two-four wire conversion module uses an impedance conversion formula: Z in =Z out ×(N1 / N2) 2 ; where Z in is the input impedance, Z out is the output impedance, and N1 / N2 is the turns ratio of the transformer.

4. The method of claim 1, wherein the method further comprises: The automatic speech recognition algorithm used by the voice recognition control module also includes a language model based on a deep neural network, and the vocabulary prediction probability formula is: ； where ω i is the current vocabulary, ω1, ω2,..., ω i-1 is the history vocabulary sequence, h i is the hidden state, and W and b are model parameters.

5. The method of claim 1, wherein the method further comprises: The VOIP analog-digital signal conversion module uses G.711 encoding protocol for analog-digital conversion, and the quantization formula is: ； wherein x is the input analog signal amplitude, y is the output digital signal, μ is the compression parameter, and is usually 255.

6. The method of claim 1, wherein the method further comprises: The core control module dynamically controls the signal paths of each module through a level signal, and the control logic uses a finite state machine model, and the state transition function is represented as: δ:S×Σ→S; wherein S is a state set, Σ is an input alphabet, and δ is a state transition function.

7. The method of claim 1, wherein the method further comprises: The method further comprises an emergency linkage processing step: when a fire alarm gateway triggers an emergency signal, the fusion communication gateway synchronously captures IP side signaling, converts dispatching instructions into analog signals, and triggers analog side talk stations and loudspeakers to respond.

8. The method of claim 6, wherein the method further comprises: The emergency linkage processing uses a priority scheduling algorithm, and the priority allocation function is: Priority = a x T emergency + b x S importance ; where T emergency is an emergency parameter, S importance is a signal importance parameter, and α and β are weight coefficients.

9. The method of claim 1, wherein the method further comprises: The VOIP analog-digital signal conversion module supports a SIP protocol stack, and the signaling packaging format complies with the RFC3261 standard, and the session establishment time satisfies: T setup ≤ 150 ms; where T setup is the total delay from signaling to session setup.

10. The method of claim 1, wherein the method further comprises: The recognition accuracy of the voice recognition control module in a noisy environment satisfies: Accuracy≥95%@SNR≥15dB; wherein SNR is the signal-to-noise ratio, and Accuracy is the voice command recognition accuracy.

11. The method of claim 1, wherein the method further comprises: The fusion communication gateway supports multi-concurrent processing, and the maximum number of concurrent channels N is determined by the following formula: ； Wherein, T processing is the processor processing capacity, T frame is the processing time of each frame, C overhead is the system overhead coefficient.

Images