Plant active and passive combined noise reduction system based on artificial intelligence and implementation method

Abstract

The invention discloses a plant active and passive combined noise reduction system based on artificial intelligence and an implementation method, and relates to the technical field of noise control, the system comprises a passive noise reduction system and a self-adaptive active noise reduction control system based on artificial intelligence; the active noise reduction control system comprises a plurality of fixed ANC active noise reducers arranged on the inner side wall of the plant and a plurality of portable magnetic ANC active noise reducers annularly distributed around high-noise equipment in the plant. The fixed ANC active noise reducers and the portable magnetic ANC active noise reducers are provided with 5G communication modules or wireless local area network WiFi modules, the fixed ANC active noise reducers and the portable magnetic ANC active noise reducers and the control host jointly form an information collection and execution module of the active noise reduction artificial intelligence large model, and the active noise reduction artificial intelligence large model collects residual noise signals in a distributed mode. The overall noise sound field environment and the noise reduction effect of each active noise reduction device are analyzed in real time, a correction scheme is made and executed autonomously, and the all-around self-adaptive active noise reduction effect is achieved.

Description

Technical Field

[0001] This invention relates to the field of noise control technology, and more specifically to the field of artificial intelligence-based active and passive combined noise reduction systems and implementation methods for factory buildings. Background Technology

[0002] With the development of modern industrial production technology, production plants are becoming increasingly larger, and workshops are filled with more and more equipment. The noise generated by this equipment is becoming increasingly difficult to control. Prolonged exposure to high-noise working environments can seriously damage the physical and mental health of workers. In modern urban life, large numbers of people live in concentrated areas, and urban development has often resulted in the juxtaposition of factory and residential areas. Consequently, noise emissions from production plants also pose a certain degree of harm to the external environment.

[0003] However, due to the great difficulty in controlling it, noise pollution has become one of the major sources of pollution in today's society, second only to air pollution, and the situation is quite serious.

[0004] Because complaints about noise pollution at factory boundaries can affect business operations, industrial enterprises generally place greater emphasis on reducing noise at factory boundaries. To reduce the pollution of the surrounding environment by factory boundary noise, passive noise reduction technologies have been widely used, such as: setting up sound barriers at factory boundaries, constructing fully enclosed factory buildings, and using soundproof windows, double-glazed windows, and laying sound-absorbing materials on the factory building envelope.

[0005] However, traditional passive noise reduction technology is only effective in enclosed production facilities. Many production processes cannot guarantee that the facility will always be enclosed, and enclosed facilities can also lead to poor air circulation and secondary air pollution. Furthermore, passive noise reduction technology is only effective at isolating high-frequency noise, but it cannot effectively isolate the mid- and low-frequency noise commonly encountered in equipment operation.

[0006] More seriously, passive noise reduction technology is ineffective inside factory buildings. For example, the operating characteristics of many pieces of equipment make it impossible to shield them with sound barriers. Consequently, there is a lack of effective means to protect workers inside the workshop, relying heavily on personal protective equipment (PPE), which is often inadequate. Long-term noise pollution can cause hearing loss, increased cardiovascular disease, and more negative emotions among workers, seriously affecting their physical and mental health and negatively impacting social productivity.

[0007] The concept of ANC (Active Noise Cancellation) technology originated in 1936, first proposed and patented by German physicist Lueg. Its technical basis lies in the physical principle that sound propagates in the air as waves, and the principle that when a pair of sound waves with the same frequency and amplitude but opposite phase are superimposed, they will cancel each other out due to interference. A pre-amplifier picks up ambient noise, and after the ANC noise reduction processing chip analyzes the noise curve, it generates a sound wave that is completely opposite in phase to the noise curve. This sound wave is then emitted by a speaker, thus canceling out the ambient noise and achieving noise reduction.

[0008] According to research, active noise cancellation technology is very effective in reducing low- and mid-frequency noise, while passive noise cancellation technology is relatively effective in isolating and reducing high-frequency noise.

[0009] Active noise cancellation technology is currently only widely used in headphones, small cars, and home appliances. However, in terms of indoor noise reduction in industrial production plants, due to limitations such as the huge sound field space and complex noise sources, it is still a blank area and cannot meet the urgent needs of workers and front-line managers to improve the noise level in their working environment.

[0010] After careful and in-depth research on the sources of noise and the ambient sound field in the production plant, the main sources of noise include: mechanical noise caused by mechanical impact, friction, and rotation from machine tools, mills, etc.; fluid dynamic noise from air compressors, fans, etc.; electromagnetic noise from generators, transformers, etc.; low-frequency vibration noise from large equipment; and human voices from production personnel. Production plant noise generally exhibits characteristics of multiple sources and mixed frequencies. The noise environment within the production plant frequently changes due to changes in production conditions (such as the addition, removal, or relocation of noisy equipment), and individual active noise reduction devices and manual control may not be able to respond in a timely manner. Equipment noise is attenuated when transmitted to active noise reduction devices in large open spaces, and the resulting data deviation affects the noise reduction effect. Different work surfaces within the plant also have different noise reduction requirements under different operating conditions, necessitating timely and precise control of active noise reduction devices.

[0011] Some existing patents can prove the above argument. For example, a patent with publication number CN217949397U discloses a new structure industrial plant noise reduction device, the contents of which are as follows: The new structure industrial plant noise reduction device includes an industrial plant body, an industrial plant burial base, a ventilation mechanism, a transparent window, a first exhaust fan, and a second exhaust fan. The middle of the bottom surface of the industrial plant body is welded to the top of the industrial plant burial base, and the middle of the upper surface of the industrial plant body is fixedly connected to the bottom surface of the ventilation mechanism by bolts. The first exhaust fan and the second exhaust fan are respectively nested in the middle of the front end and the top of the front of the industrial plant body. The main body of the industrial plant includes an outer chamber, a door mounting hole, an inner chamber, sound-absorbing panels, a cavity, and a rectangular mounting opening. The door mounting hole is located at the lower end of one end of the outer chamber's surface, and the inner chamber is nested within the middle of the outer chamber. A cavity exists between the outer and inner chambers, and sound-absorbing panels are symmetrically installed on the inner surface of the outer chamber. A rectangular mounting opening is located in the middle of the bottom surface of the outer chamber. In use, all noise generated inside the inner chamber is concentrated and isolated within the cavity, and silenced by the sound-absorbing panels. Combined with the first and second exhaust fans, the exhaust air and noise are further concentrated and silenced. This makes the entire structure act like a large silencer, effectively reducing noise not only from machinery but also from other sources, thus suppressing noise transmission and minimizing its impact on the outside environment.

[0012] However, this patent has significant shortcomings. The main issue is that its noise reduction function only considers "effectively suppressing noise transmission and reducing its impact on the outside world." It completely neglects the equally important issue of noise control within the factory premises, resulting in a clear flaw in the solution.

[0013] For example, a patent with publication number CN208633301U discloses a novel noise reduction device for industrial plants, the contents of which are as follows: A novel noise reduction device for industrial plants includes an industrial plant body, a noise reduction device, a concrete floor, and a vibration damping device. The industrial plant body is installed on the upper surface of the concrete floor. An installation groove is provided at the bottom of the interior of the industrial plant body. The installation groove is located on the upper surface of the concrete floor. The noise reduction device is installed inside the installation groove. The noise reduction device includes an upper steel plate, a vibration damping device, and a lower steel plate. The upper steel plate and the lower steel plate are respectively located at the upper and lower ends of the vibration damping device, ...

[0014] However, this patent also has significant shortcomings. These are mainly reflected in: Although the noise reduction function of this patent takes into account "reducing the noise generated by the vibration of large machinery working inside the industrial plant", it is difficult to achieve, both technically and economically, for large plants to use the entire ground as a vibration isolation and noise reduction support.

[0015] There are numerous noise sources inside the production plant, each with its own unique frequency and propagation path. Simply reducing noise through foundation vibration without addressing other noise sources is far from sufficient. Summary of the Invention

[0016] The purpose of this invention is to solve the technical problems of active and passive noise reduction in factory buildings. This invention provides an artificial intelligence-based active and passive combined noise reduction system and implementation method for factory buildings.

[0017] To achieve the above objectives, the present invention specifically adopts the following technical solution: One aspect of the present invention provides an artificial intelligence-based active and passive combined noise reduction system for factory buildings, including a passive noise reduction system and an artificial intelligence-based adaptive active noise reduction control system; The active noise cancellation control system includes several fixed ANC active noise cancellers installed on the inner wall of the factory and several portable magnetic ANC active noise cancellers distributed in a ring around the high-noise equipment in the factory. Each fixed ANC active noise canceller and each portable magnetic ANC active noise canceller is equipped with a 5G communication module or a wireless local area network (WiFi) module. Together with the control host (one unit or several units in a hierarchical manner), they form the information collection and execution module of the active noise cancellation artificial intelligence big model. The active noise cancellation artificial intelligence big model collects residual noise signals in a distributed manner, analyzes the overall noise sound field environment and the noise reduction effect of each active noise canceller in real time, autonomously formulates and executes correction schemes, and achieves a comprehensive adaptive active noise cancellation effect.

[0018] Specifically, to meet the noise reduction requirements of complex sound field environments, this solution designs two types of ANC (Active Noise Cancellation) devices. Both devices perform active noise cancellation and employ the same noise reduction technology, but differ in installation location, installation method, device power, and device shape. To achieve better active noise cancellation functionality, both ANC devices utilize a feedforward feedback hybrid ANC technology.

[0019] In one embodiment, several fixed ANC active noise cancellers are divided into multiple groups, and the multiple groups of fixed ANC active noise cancellers are distributed in an array on the inner wall of the factory.

[0020] Specifically, several fixed ANC active noise cancellers are suspended from the upper part of the factory columns, forming a uniformly distributed array that covers the entire factory interior, reducing the overall environmental noise inside the factory.

[0021] In one embodiment, each fixed ANC active noise canceller includes a housing, a pickup unit, a speaker unit, a PCB unit, and a back cover; The pickup unit, speaker unit, and PCB unit are encapsulated in a housing formed by the encapsulation shell and the back cover, and the pickup unit, speaker unit, and PCB unit are all located on the same plane; Multiple holes covered with acoustically transparent fabric are opened on the front panel of the package housing. Some holes are used to collect ambient noise, while others are used to allow anti-noise sound waves to pass through. The speaker unit includes at least two high-frequency speakers and one mid-low frequency speaker; The pickup unit includes at least two front ambient noise acquisition modules and at least two residual noise acquisition modules; The PCB unit integrates an ANC active noise cancellation module, an audio amplifier module, a power supply module, a 5G communication module, and a wireless local area network (WiFi) module.

[0022] Specifically, the package used as an example in this solution is designed as a cuboid. Multiple holes covered with acoustically transparent fabric are present on the front panel of the package; some are used to collect ambient noise, while others allow noise-reflecting sound waves to pass through. Both the high-frequency and mid-low-frequency speakers are omnidirectional. The fixed ANC active noise canceller is directly powered by AC, which is then converted to DC before operating.

[0023] A communication module can be integrated into the PCB design of a fixed ANC active noise cancelling unit to connect with the factory broadcasting system. This allows for functions such as voice announcements and background music playback, controlled by the broadcasting system. Background music helps further reduce the adverse psychological effects of noise.

[0024] In one embodiment, several portable magnetic ANC active noise cancelling devices are magnetically installed on the surface of high-noise equipment and distributed in a ring around the high-noise equipment. The installation height of the portable magnetic ANC active noise cancelling devices is 1.5-1.7m above the ground, and the sound field of the portable magnetic ANC active noise cancelling devices covers the working range of the operator.

[0025] In one embodiment, the portable magnetic ANC active noise canceller includes a portable housing, a portable pickup unit, a portable speaker unit, a portable PCB unit, and a portable back cover. The portable pickup unit, portable speaker unit, and portable PCB unit are packaged in a portable housing formed by the portable housing and the portable back cover. The portable pickup unit and the portable speaker unit are located on the same plane. The portable speaker unit uses an omnidirectional speaker; The portable pickup unit includes multiple front-end ambient noise acquisition modules and at least two residual noise acquisition modules; The portable PCB unit integrates a portable ANC active noise cancellation module, a portable audio amplifier module, a portable power supply module, a portable 5G communication module, and a portable wireless local area network (WiFi) module. The portable magnetic ANC active noise cancelling device has a built-in rechargeable battery and a Type-C USB universal charging port on its portable casing.

[0026] Specifically, the portable packaging shell used as an example of this solution is designed to be circular. Holes covered with sound-permeable fabric are opened on the front panel of the portable packaging shell and the surrounding annular shell, which is more conducive to collecting ambient noise and to allowing anti-noise sound waves to pass through.

[0027] In one embodiment, the passive noise reduction system includes a fully enclosed building enclosure, perforated sound-absorbing and sound-insulating composite panels installed in the building walls, soundproof doors installed in the building walls, soundproof windows or double-glazed windows installed in the building walls, and sound-absorbing material layers laid in ventilation devices or other openings in the building walls. Passive noise reduction systems also include soundproof enclosures that partially enclose the exterior of high-noise equipment, vibration damping supports or ground isolation installed under the foundations of equipment with severe vibration.

[0028] Specifically, the above measures can absorb and reduce some of the high-frequency and low-frequency noise, thus achieving passive noise reduction in the factory and some equipment.

[0029] In addition, there is a type of large-scale special equipment whose noise is characterized by strong low-frequency vibrations during operation. Not only does the equipment itself emit low-frequency noise, but surrounding buildings and floors also receive and propagate these vibrations, generating noise as well. This type of high-intensity low-frequency noise can damage hearing and also harm the nervous, cardiovascular, and endocrine systems. To address this type of low-frequency noise from vibrating equipment, this invention also provides active vibration countermeasures technology as an important technical component.

[0030] Vibration testing is conducted on the equipment before it leaves the factory to measure its vibration displacement, velocity, and acceleration. After analysis, parameters such as vibration frequency, phase, and impact value are generated.

[0031] Design an active anti-phase vibration generator based on vibration parameters.

[0032] When installing vibration equipment, the equipment foundation should be isolated from the surrounding ground, and an active anti-phase vibration generator should be placed in the isolation gap.

[0033] When the vibrating equipment is running, the active anti-phase vibration generator is started synchronously, and the working state of the generator is finely adjusted so that the two vibrations are in the same frequency but opposite phase and cancel each other out.

[0034] Another aspect of the present invention provides an artificial intelligence-based method for combined active and passive noise reduction in factory buildings, employing the aforementioned artificial intelligence-based combined active and passive noise reduction system for factory buildings, comprising the following steps: S1. Selection of active noise reduction AI large model: Select a multimodal AI large model that can accept sound input and understands the laws of acoustic physics. Currently, the basic large models available are Kimi-Audio and MiDashengLM, which are AI large models with noise resistance testing and environmental sound understanding capabilities. S2, Pre-training of large-scale AI models for active noise reduction; S3. Fine-tuning of the active noise cancellation AI model: Deploy the active noise cancellation AI model in the actual environment of an industrial plant. Obtain data such as the location of various noise sources, noise spectrum, noise loudness, noise attenuation, distance between noise pickup units and noise sources, and the relative positions of the inspection personnel's work surface, noise sources, and anti-noise loudspeakers through noise pickup units distributed throughout the plant. Calculate the spectrum and loudness of the anti-noise output by the anti-noise loudspeakers, output anti-noise through the anti-noise loudspeakers, and detect the noise reduction effect. Fine-tune the model based on the detected noise reduction effect. S4. Application of the active noise cancellation AI model: Deploy the active noise cancellation AI model in the actual factory environment, coordinate and control the output of each fixed ANC active noise canceller to counter noise, and use adaptive filtering algorithms (such as FxLMS) to achieve millisecond-level response and carry out precise active noise reduction operations on the work surface. S5. Timely updates of fixed ANC active noise cancellation devices: When the actual environment in the industrial plant changes, such as changes in the noise source, intensity, or location, changes in the spatial location of the work surface, or changes in the location of the noise-reducing loudspeakers, the large model will autonomously and promptly readjust, generate new noise reduction strategies, and execute them.

[0035] In one implementation, the specific method for pre-training the active denoising AI large model in step S2 is as follows: S21. Sound Field Perception: Train an active noise reduction AI model to identify various noise sources in the factory environment, the causes of noise generation, noise frequency, noise loudness, noise duration, noise attenuation, and other data (the noise data used for training can be obtained from public noise databases or collected on-site in the application environment and generated through analog-to-digital conversion), and use CNN algorithms to identify the frequency characteristics of noise sources. S22, Inverse acoustic wave data generation: Use a deep learning model (such as LSTM) to predict noise changes, and train an active noise reduction artificial intelligence large model based on the noise data to generate anti-noise data that is in the opposite phase to the noise waveform; S23. Dynamic Control: Using reinforcement learning algorithms (such as DQN), the system continuously optimizes speaker output parameters through digital-to-analog conversion and noise reduction of the speaker output, adapting to changes in equipment operating conditions. A MIMO control system is used to address the distortion caused by superposition of sound waves in multi-reflection environments. In one implementation, step S5 includes changes in the actual environment inside the industrial plant, including changes in the noise source (increase or decrease), changes in the intensity of the noise source, changes in the location of the noise source, changes in the spatial location of the work surface, and changes in the location of the anti-noise loudspeaker.

[0036] In one implementation, the workflow of the fixed ANC active noise cancellation device in step S4 is as follows: S41. The first active noise cancellation operation is completed by each individual fixed ANC active noise canceller in the noise cancellation control network according to the pre-trained instructions generated by the human active noise cancellation artificial intelligence model. S42. Each individual fixed ANC active noise canceller completes its work according to the initial instructions, while collecting residual noise through the built-in residual noise collection microphone, performing analog-to-digital conversion (ADC) using the integrated chip, and transmitting the digital signal to the control host through the communication network. S43. The active noise cancellation artificial intelligence big model running in the control host receives the digital signals transmitted by each individual fixed ANC active noise cancellation unit, analyzes the noise sound field environment, evaluates the noise reduction effect, formulates a correction plan after calculation, adjusts the spectrum, delay and output intensity data of the anti-noise output of some noise cancellation units in the array, and transmits the corresponding digital signals to the individual fixed ANC active noise cancellation units that need to be adjusted for execution. S44. Repeat the above system process, and perform multiple rounds of optimization through the deep learning algorithm of the active noise cancellation artificial intelligence large model to form a stable noise cancellation scheme within a certain period of time and hand it over to each individual fixed ANC active noise cancellation unit for execution. S45. When the noise environment inside the factory changes significantly and the residual noise after noise reduction exceeds the limit threshold, the active noise reduction AI big data model automatically returns to execute a new learning and decision-making process, enters a new cycle, until a new stable noise reduction scheme is formed and executed.

[0037] Specifically, the above system process can achieve better, more comprehensive, adaptive active noise reduction. After the system has been widely adopted and a certain amount of data has been collected, it can be combined with 5G networks, cloud computing, and big data technologies to conduct systematic data analysis on various types of high-noise equipment and various types of production plants, generating more adaptable pre-built expert solutions that can be promoted to other plants to achieve even better active noise reduction.

[0038] When both types of ANC (Active Noise Cancellation) are in operation, the pre-amplifier ambient noise acquisition module in the pickup unit first collects omnidirectional ambient noise. Then, the ANC processing chip integrated in the PCB unit performs analog-to-digital conversion (ADC) and digital analysis (DSP) to generate a noise curve. This information is transmitted to the active noise cancellation AI central control system, which calculates and generates inverted sound wave data. This data is then sent back to the ANC active noise canceller and converted to analog-to-digital conversion (DAC) before being emitted as "anti-noise" through the speaker, thus realizing the feedforward ANC process. Simultaneously, the residual noise acquisition module in the pickup unit collects the residual noise after noise reduction processing and sends it back to the ANC processing chip for further processing, realizing the feedback ANC process. This hybrid feedback ANC technology can achieve better noise reduction results.

[0039] The beneficial effects of this invention are as follows: 1. This invention differs from traditional factory noise reduction technologies that rely solely on passive methods such as sound insulation and absorption. It provides an AI-based system control method that, through corresponding active noise reduction equipment, achieves an adaptive, comprehensive, and proactive noise reduction solution for factories. Furthermore, it uses traditional passive noise reduction technologies as supplementary means to achieve overall efficiency improvement. While further enhancing the overall noise reduction level, it overcomes the shortcomings of traditional methods that can only suppress noise leakage to the outside of the factory but cannot effectively reduce noise inside the factory. This not only reduces noise pollution from the production plant to the surrounding environment but also better protects the physical and mental health of workers inside the factory.

[0040] 2. The solution covers system control methods based on artificial intelligence, a solution that combines active and passive noise reduction technologies, the design of two types of ANC active noise reducers for different locations, and the technology of using an active anti-phase vibration generator to solve low-frequency vibration noise of equipment.

[0041] 3. This invention can truly and effectively solve the noise reduction problem in production plants, reduce the adverse effects of noise on the external environment and production personnel inside the plant, and is more conducive to maintaining people's physical and mental health, improving people's living and working conditions, and thus contributing to maintaining social productivity. Attached Figure Description

[0042] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained from these drawings without creative effort.

[0043] Figure 1 This is a schematic diagram of the ANC (Active Noise Cancellation) principle.

[0044] Figure 2 This is a schematic diagram showing the distribution of noise sources and active / passive noise reduction systems in the production plant.

[0045] Figure 3 This is a schematic diagram of a combined active and passive noise reduction system for a production plant.

[0046] Figure 4 This is a flowchart of a feedforward / feedback hybrid ANC system.

[0047] Figure 5 This is an exploded view of the portable magnetic ANC active noise cancellation device.

[0048] Figure 6 This is a schematic diagram of the assembly of a portable magnetic ANC active noise cancelling device.

[0049] Figure 7 This is an exploded view of a fixed ANC active noise canceller.

[0050] Figure 8 This is a schematic diagram of a fixed ANC active noise cancellation unit assembly.

[0051] Figure 9 This is a flowchart illustrating the deployment and usage of a large-scale AI model for active noise reduction in factory buildings.

[0052] Figure 10 This is a control flowchart for an active noise reduction control system based on the Internet of Things and deep learning.

[0053] Figure 11 This is a layout diagram of an active anti-phase vibration system. Detailed Implementation

[0054] To make the technical problems, technical solutions, and technical effects of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0055] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0056] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0057] In the description of the embodiments of the present invention, it should be noted that the terms "inner", "outer", "upper", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship in which the product of the invention is usually placed when in use. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting the present invention.

[0058] Example 1 like Figures 1 to 11 As shown, this embodiment provides an artificial intelligence-based active and passive noise reduction system for factory buildings, including a passive noise reduction system and an artificial intelligence-based adaptive active noise reduction control system. The active noise cancellation control system includes several fixed ANC active noise cancellers installed on the inner wall of the factory and several portable magnetic ANC active noise cancellers distributed in a ring around the high-noise equipment in the factory. Each fixed ANC active noise canceller and each portable magnetic ANC active noise canceller is equipped with a 5G communication module or a wireless local area network (WiFi) module. Together with the control host (one unit or several units in a hierarchical manner), they form the information collection and execution module of the active noise cancellation artificial intelligence big model. The active noise cancellation artificial intelligence big model collects residual noise signals in a distributed manner, analyzes the overall noise sound field environment and the noise reduction effect of each active noise canceller in real time, autonomously formulates and executes correction schemes, and achieves a comprehensive adaptive active noise cancellation effect.

[0059] Specifically, to meet the noise reduction requirements of complex sound field environments, this solution designs two types of ANC (Active Noise Cancellation) devices. Both devices perform active noise cancellation and employ the same noise reduction technology, but differ in installation location, installation method, device power, and device shape. To achieve better active noise cancellation functionality, both ANC devices utilize a feedforward feedback hybrid ANC technology.

[0060] Example 2 like Figure 7 and Figure 8 As shown, this embodiment provides an artificial intelligence-based active and passive noise reduction system for factory buildings, characterized in that it includes a passive noise reduction system and an artificial intelligence-based adaptive active noise reduction control system. The active noise cancellation control system includes several fixed ANC active noise cancellers installed on the inner wall of the factory and several portable magnetic ANC active noise cancellers distributed in a ring around the high-noise equipment in the factory. Each fixed ANC active noise canceller and each portable magnetic ANC active noise canceller is equipped with a 5G communication module or a wireless local area network (WiFi) module. Together with the control host (one unit or several units in a hierarchical manner), they form an active noise cancellation artificial intelligence large model. The active noise cancellation artificial intelligence large model collects residual noise signals in a distributed manner, analyzes the overall noise sound field environment and the noise reduction effect of each active noise canceller in real time, autonomously formulates and executes correction schemes, and achieves a comprehensive adaptive active noise cancellation effect.

[0061] Several fixed ANC active noise cancellers are divided into multiple groups, and the multiple groups of fixed ANC active noise cancellers are distributed in an array on the inner wall of the factory.

[0062] Specifically, several fixed ANC active noise cancellers are suspended from the upper part of the factory columns, forming a uniformly distributed array that covers the entire factory interior, reducing the overall environmental noise inside the factory.

[0063] Each fixed ANC active noise canceller includes a housing, a pickup unit, a speaker unit, a PCB unit, and a back cover; The pickup unit, speaker unit, and PCB unit are encapsulated in a housing formed by the encapsulation shell and the back cover, and the pickup unit, speaker unit, and PCB unit are all located on the same plane; Multiple holes covered with acoustically transparent fabric are opened on the front panel of the package housing. Some holes are used to collect ambient noise, while others are used to allow anti-noise sound waves to pass through. The speaker unit includes at least two high-frequency speakers and one mid-low frequency speaker; The pickup unit includes at least two front ambient noise acquisition modules and at least two residual noise acquisition modules; The PCB unit integrates an ANC active noise cancellation module, an audio amplifier module, a power supply module, a 5G communication module, and a wireless local area network (WiFi) module.

[0064] Specifically, the package used as an example in this solution is designed as a cuboid. Multiple holes covered with acoustically transparent fabric are present on the front panel of the package; some are used to collect ambient noise, while others allow noise-reflecting sound waves to pass through. Both the high-frequency and mid-low-frequency speakers are omnidirectional. The fixed ANC active noise canceller is directly powered by AC, which is then converted to DC before operating.

[0065] A communication module can be integrated into the PCB design of a fixed ANC active noise cancelling unit to connect with the factory broadcasting system. This allows for functions such as voice announcements and background music playback, controlled by the broadcasting system. Background music helps further reduce the adverse psychological effects of noise.

[0066] Example 3 like Figure 5 and Figure 6 As shown, this embodiment is a further optimization based on embodiment 2, as detailed below: Several portable magnetic ANC active noise cancelling devices are magnetically installed on the surface of high-noise equipment and distributed in a ring around the high-noise equipment. The installation height of the portable magnetic ANC active noise cancelling devices is 1.5-1.7m above the ground, and the sound field of the portable magnetic ANC active noise cancelling devices covers the working range of the operator.

[0067] In one embodiment, the portable magnetic ANC active noise canceller includes a portable housing, a portable pickup unit, a portable speaker unit, a portable PCB unit, and a portable back cover. The portable pickup unit, portable speaker unit, and portable PCB unit are packaged in a portable housing formed by the portable housing and the portable back cover. The portable pickup unit and the portable speaker unit are located on the same plane. The portable speaker unit uses an omnidirectional speaker; The portable pickup unit includes multiple front-end ambient noise acquisition modules and at least two residual noise acquisition modules; The portable PCB unit integrates a portable ANC active noise cancellation module, a portable audio amplifier module, a portable power supply module, a portable 5G communication module, and a portable wireless local area network (WiFi) module. The portable magnetic ANC active noise cancelling device has a built-in rechargeable battery and a Type-C USB universal charging port on its portable casing.

[0068] Specifically, the portable packaging shell used as an example of this solution is designed to be circular. Holes covered with sound-permeable fabric are opened on the front panel of the portable packaging shell and the surrounding annular shell, which is more conducive to collecting ambient noise and to allowing anti-noise sound waves to pass through.

[0069] Example 4 like Figure 3 As shown, this embodiment is a further optimization based on embodiment 2, as detailed below: In one embodiment, the passive noise reduction system includes a fully enclosed building enclosure, perforated sound-absorbing and sound-insulating composite panels installed in the building walls, soundproof doors installed in the building walls, soundproof windows or double-glazed windows installed in the building walls, and sound-absorbing material layers laid in ventilation devices or other openings in the building walls. Passive noise reduction systems also include soundproof enclosures that partially enclose the exterior of high-noise equipment, vibration damping supports or ground isolation installed under the foundations of equipment with severe vibration.

[0070] Specifically, the above measures can absorb and reduce some of the high-frequency and low-frequency noise, thus achieving passive noise reduction in the factory and some equipment.

[0071] In addition, there is a type of large-scale special equipment whose noise is characterized by strong low-frequency vibrations during operation. Not only does the equipment itself emit low-frequency noise, but surrounding buildings and floors also receive and propagate these vibrations, generating noise as well. This type of high-intensity low-frequency noise can damage hearing and also harm the nervous, cardiovascular, and endocrine systems. To address this type of low-frequency noise from vibrating equipment, this invention also provides active vibration countermeasures technology as an important technical component.

[0072] Vibration testing is conducted on the equipment before it leaves the factory to measure its vibration displacement, velocity, and acceleration. After analysis, parameters such as vibration frequency, phase, and impact value are generated.

[0073] Design an active anti-phase vibration generator based on vibration parameters.

[0074] When installing vibration equipment, the equipment foundation should be isolated from the surrounding ground, and an active anti-phase vibration generator should be placed in the isolation gap.

[0075] When the vibrating equipment is running, the active anti-phase vibration generator is started synchronously, and the working state of the generator is finely adjusted so that the two vibrations are in the same frequency but opposite phase and cancel each other out.

[0076] Example 5 This embodiment provides a method for achieving combined active and passive noise reduction in factory buildings based on artificial intelligence, including the following steps: S1. Selection of active noise reduction AI large model: Select a multimodal AI large model that can accept sound input and understands the laws of acoustic physics. Currently, the basic large models available are Kimi-Audio and MiDashengLM, which are AI large models with noise resistance testing and environmental sound understanding capabilities. S2. Pre-training of large-scale AI models for active noise reduction: S21. Sound Field Perception: Train an active noise reduction AI model to identify various noise sources in the factory environment, the causes of noise generation, noise frequency, noise loudness, noise duration, noise attenuation, and other data (the noise data used for training can be obtained from public noise databases or collected on-site in the application environment and generated through analog-to-digital conversion), and use CNN algorithms to identify the frequency characteristics of noise sources. S22, Inverse acoustic wave data generation: Use a deep learning model (such as LSTM) to predict noise changes, and train an active noise reduction artificial intelligence large model based on the noise data to generate anti-noise data that is in the opposite phase to the noise waveform; S23. Dynamic Control: Using reinforcement learning algorithms (such as DQN), the system continuously optimizes speaker output parameters through digital-to-analog conversion and noise reduction of the speaker output, adapting to changes in equipment operating conditions. A MIMO control system is used to address the distortion caused by superposition of sound waves in multi-reflection environments. S3. Fine-tuning of the active noise cancellation AI model: Deploy the active noise cancellation AI model in the actual environment of an industrial plant. Obtain data such as the location of various noise sources, noise spectrum, noise loudness, noise attenuation, distance between noise pickup units and noise sources, and the relative positions of the inspection personnel's work surface, noise sources, and anti-noise loudspeakers through noise pickup units distributed throughout the plant. Calculate the spectrum and loudness of the anti-noise output by the anti-noise loudspeakers, output anti-noise through the anti-noise loudspeakers, and detect the noise reduction effect. Fine-tune the model based on the detected noise reduction effect. S4. Application of the Active Noise Cancellation AI Model: The active noise cancellation AI model is deployed in the actual factory environment to coordinate and control the output noise reduction of each fixed ANC active noise canceller. Adaptive filtering algorithms (such as FxLMS) are used to achieve millisecond-level response, enabling precise active noise reduction operations on the work surface. The workflow of the fixed ANC active noise canceller is as follows: S41. The first active noise cancellation operation is completed by each individual fixed ANC active noise canceller in the noise cancellation control network according to the pre-trained instructions generated by the human active noise cancellation artificial intelligence model. S42. Each individual fixed ANC active noise canceller completes its work according to the initial instructions, while collecting residual noise through the built-in residual noise collection microphone, performing analog-to-digital conversion (ADC) using the integrated chip, and transmitting the digital signal to the control host through the communication network. S43. The active noise cancellation artificial intelligence big model running in the control host receives the digital signals transmitted by each individual fixed ANC active noise cancellation unit, analyzes the noise sound field environment, evaluates the noise reduction effect, formulates a correction plan after calculation, adjusts the spectrum, delay and output intensity data of the anti-noise output of some noise cancellation units in the array, and transmits the corresponding digital signals to the individual fixed ANC active noise cancellation units that need to be adjusted for execution. S44. Repeat the above system process, and perform multiple rounds of optimization through the deep learning algorithm of the active noise cancellation artificial intelligence large model to form a stable noise cancellation scheme within a certain period of time and hand it over to each individual fixed ANC active noise cancellation unit for execution. S45. When the noise environment inside the factory changes significantly and the residual noise after noise reduction exceeds the limit threshold, the active noise reduction AI big data model automatically returns to execute a new learning and decision-making process, enters a new cycle, until a new stable noise reduction scheme is formed and executed.

[0077] S5. Timely updates of fixed ANC active noise cancellation devices: When the actual environment in the industrial plant changes, such as changes in the noise source, intensity, or location, changes in the spatial location of the work surface, or changes in the location of the noise-reducing loudspeakers, the large model will autonomously and promptly readjust, generate new noise reduction strategies, and execute them.

[0078] In step S5, changes in the actual environment inside the industrial plant include changes in the number of noise sources, changes in the intensity of noise sources, changes in the location of noise sources, changes in the spatial location of the work surface, and changes in the location of the noise-reducing loudspeakers.

[0079] In this embodiment, the above system process achieves a better, more comprehensive, adaptive active noise reduction effect. After the system is widely adopted and a certain amount of data is obtained, it can be combined with 5G networks, cloud computing, and big data technologies to conduct systematic data analysis on various types of high-noise equipment and various types of production plants, generating more adaptable pre-set expert solutions that can be promoted to other plants to achieve even better active noise reduction effects.

[0080] When both types of ANC (Active Noise Cancellation) are in operation, the pre-amplifier ambient noise acquisition module in the pickup unit first collects omnidirectional ambient noise. Then, the ANC processing chip integrated in the PCB unit performs analog-to-digital conversion (ADC) and digital analysis (DSP) to generate a noise curve. This information is transmitted to the active noise cancellation AI central control system, which calculates and generates inverted sound wave data. This data is then sent back to the ANC active noise canceller and converted to analog-to-digital conversion (DAC) before being emitted as "anti-noise" through the speaker, thus realizing the feedforward ANC process. Simultaneously, the residual noise acquisition module in the pickup unit collects the residual noise after noise reduction processing and sends it back to the ANC processing chip for further processing, realizing the feedback ANC process. This feedforward and feedback hybrid ANC technology can achieve better noise reduction results.

Claims

1. An artificial intelligence-based active and passive combined noise reduction system for factory buildings, characterized in that: This includes passive noise reduction systems and AI-based adaptive active noise reduction control systems; The active noise reduction control system includes several fixed ANC active noise reducers installed on the inner wall of the factory building and several portable magnetic ANC active noise reducers distributed in a ring around the high-noise equipment in the factory building. Each of the fixed ANC active noise cancellers and each of the portable magnetic ANC active noise cancellers are equipped with a 5G communication module or a wireless local area network (WiFi) module. Together with the control host, they form the information collection and execution module of the active noise cancellation artificial intelligence big model. The active noise cancellation artificial intelligence big model collects residual noise signals in a distributed manner, analyzes the overall noise sound field environment and the noise reduction effect of each active noise canceller in real time, autonomously formulates and executes correction schemes, and achieves a comprehensive adaptive active noise cancellation effect.

2. The artificial intelligence-based active and passive combined noise reduction system for factory buildings according to claim 1, characterized in that, The fixed ANC active noise cancellers are divided into multiple groups, and the multiple groups of fixed ANC active noise cancellers are distributed in an array on the inner wall of the factory.

3. The artificial intelligence-based active and passive combined noise reduction system for factory buildings according to claim 2, characterized in that, Each of the aforementioned fixed ANC active noise cancelling units includes a package housing, a pickup unit, a speaker unit, a PCB unit, and a rear cover; The pickup unit, the speaker unit, and the PCB unit are encapsulated in a housing formed by the encapsulation shell and the back cover, and the pickup unit, the speaker unit, and the PCB unit are all located on the same plane; Multiple holes covered with sound-permeable fabric are opened on the front panel of the encapsulation shell. Some of the holes are used to collect ambient noise, while others are used to allow anti-noise sound waves to pass through. The loudspeaker unit includes at least two high-frequency loudspeakers and one mid-low frequency loudspeaker; The pickup unit includes at least two front-end ambient noise acquisition modules and at least two residual noise acquisition modules; The PCB unit integrates an ANC active noise cancellation module, an audio amplifier module, a power supply module, a 5G communication module, and a WiFi module.

4. The artificial intelligence-based active and passive combined noise reduction system for factory buildings according to claim 3, characterized in that, Several portable magnetic ANC active noise cancelling devices are magnetically installed on the surface of high-noise equipment and distributed in a ring around the high-noise equipment. The installation height of several portable magnetic ANC active noise cancelling devices is 1.5-1.7m above the ground, and the sound field of several portable magnetic ANC active noise cancelling devices covers the working range of the operator.

5. The artificial intelligence-based active and passive combined noise reduction system for factory buildings according to claim 4, characterized in that, The portable magnetic ANC active noise cancelling device includes a portable housing, a portable pickup unit, a portable speaker unit, a portable PCB unit, and a portable back cover. The portable pickup unit, the portable speaker unit, and the portable PCB unit are encapsulated in the portable housing formed by the portable housing and the portable back cover. The portable pickup unit and the portable speaker unit are located on the same plane. The portable speaker unit is an omnidirectional speaker; The portable pickup unit includes multiple front-end ambient noise acquisition modules and at least two residual noise acquisition modules. The portable PCB unit integrates a portable ANC active noise cancellation module, a portable audio amplifier module, a portable power supply module, a portable 5G communication module, and a portable wireless local area network (WiFi) module. The portable magnetic ANC active noise cancelling device is equipped with a rechargeable battery, and the portable casing is equipped with a Type-C USB universal charging port.

6. The artificial intelligence-based active and passive combined noise reduction system for factory buildings according to claim 5, characterized in that, The passive noise reduction system includes a fully enclosed enclosure structure of the factory building, perforated sound-absorbing and sound-insulating composite panels installed in the factory building walls, soundproof doors installed in the factory building walls, soundproof windows or double-glazed windows installed in the factory building walls, and sound-absorbing material layers laid in the ventilation devices or other openings of the factory building walls. The passive noise reduction system also includes soundproof covers that partially enclose the high-noise equipment, vibration damping supports or ground isolation installed under the foundation of equipment with severe vibration.

7. A method for achieving combined active and passive noise reduction in factory buildings based on artificial intelligence, employing the combined active and passive noise reduction system for factory buildings based on artificial intelligence as described in claim 6, characterized in that... Includes the following steps: S1. Selection of active noise reduction AI large model: Select a multimodal AI large model that can accept sound input and understands the laws of acoustic physics. Currently, the two basic large models available are Kimi-Audio and MiDashengLM, which have noise resistance testing and environmental sound understanding capabilities. S2, Pre-training of large-scale AI models for active noise reduction; S3. Fine-tuning of the active noise cancellation AI model: Deploy the active noise cancellation AI model in the actual environment of an industrial plant. Obtain data on the location, noise spectrum, noise loudness, noise attenuation, distance between the noise pickup unit and the noise source, and the relative positions of the inspection personnel's work surface, the noise source, and the anti-noise loudspeaker through noise pickup units distributed throughout the plant. Calculate the spectrum and loudness of the anti-noise output by the anti-noise loudspeaker, output the anti-noise through the anti-noise loudspeaker, and detect the noise reduction effect. Fine-tune the model based on the detected noise reduction effect. S4. Application of the active noise cancellation AI model: Deploy the active noise cancellation AI model in the actual factory environment, coordinate and control the output of each fixed ANC active noise canceller to counter noise, and use adaptive filtering algorithms (such as FxLMS) to achieve millisecond-level response and carry out precise active noise reduction operations on the work surface. S5. Timely updates of fixed ANC active noise cancellation devices: When the actual environment in the industrial plant changes, such as changes in the noise source, intensity, or location, changes in the spatial location of the work surface, or changes in the location of the noise-reducing loudspeakers, the large model will autonomously and promptly readjust, generate new noise reduction strategies, and execute them.

8. The method for achieving combined active and passive noise reduction in factory buildings based on artificial intelligence according to claim 7, characterized in that, In step S2, the specific method for pre-training the active denoising AI large model is as follows: S21. Sound Field Perception: Train an active noise reduction AI model to identify various noise sources in the factory environment, the causes of noise generation, noise frequency, noise loudness, noise duration, and noise attenuation data, and use CNN algorithms to identify the frequency characteristics of noise sources. S22, Inverse acoustic wave data generation: Using a deep learning model to predict noise changes, an active noise reduction artificial intelligence large model is trained based on the noise data to generate inverse noise data that is inversely phase to the noise waveform; S23. Dynamic control: Using reinforcement learning algorithms, the system continuously optimizes the output parameters of the speaker by converting digital to analog and outputting anti-noise through anti-noise conversion, in order to adapt to changes in equipment operating conditions. The system also uses a MIMO control system to counteract the distortion caused by the superposition of sound waves in a multi-reflection environment.

9. The method for achieving combined active and passive noise reduction in factory buildings based on artificial intelligence according to claim 7, characterized in that, In step S5, changes in the actual environment inside the industrial plant include changes in the number of noise sources, changes in the intensity of noise sources, changes in the location of noise sources, changes in the spatial location of the work surface, and changes in the location of the noise-reducing loudspeakers.

10. The method for achieving combined active and passive noise reduction in factory buildings based on artificial intelligence according to claim 7, characterized in that, In step S4, the workflow of the fixed ANC active noise cancellation device is as follows: S41. The first active noise cancellation operation is completed by each individual fixed ANC active noise canceller in the noise cancellation control network according to the pre-trained instructions generated by the human active noise cancellation artificial intelligence model. S42. Each individual fixed ANC active noise canceller completes its work according to the initial instructions, while collecting residual noise through the built-in residual noise collection microphone, performing analog-to-digital conversion using the integrated chip, and transmitting the digital signal to the control host through the communication network. S43. The active noise cancellation artificial intelligence big model running in the control host receives the digital signals transmitted by each individual fixed ANC active noise cancellation unit, analyzes the noise sound field environment, evaluates the noise reduction effect, formulates a correction plan after calculation, adjusts the spectrum, delay and output intensity data of the anti-noise output of some noise cancellation units in the array, and transmits the corresponding digital signals to the individual fixed ANC active noise cancellation units that need to be adjusted for execution. S44. Repeat the above system process, and perform multiple rounds of optimization through the deep learning algorithm of the active noise cancellation artificial intelligence large model to form a stable noise cancellation scheme within a certain period of time and hand it over to each individual fixed ANC active noise cancellation unit for execution. S45. When the noise environment inside the factory changes significantly and the residual noise after noise reduction exceeds the limit threshold, the active noise reduction AI big data model automatically returns to execute a new learning and decision-making process, enters a new cycle, until a new stable noise reduction scheme is formed and executed.

Images