Escalator remote function safety device based on safety PLC and control method

Abstract

This invention provides a remote functional safety device and control method for escalators based on a safety PLC, relating to the field of elevator safety control technology. The remote functional safety device for escalators based on a safety PLC includes: an AI abnormal behavior detection module; a main control room monitoring and output module; a signal transmission module; an escalator control cabinet detection module; and a safety loop control module. Through AI behavior recognition, safety PLC closed-loop control, and long-distance signal transmission, it enables remote emergency stopping of the escalator under abnormal operating conditions, meeting the SIL2 safety integrity level requirements of the IEC61508 standard. It is suitable for escalators and moving walkways in densely populated places such as shopping malls, subway stations, and airports.

Description

Technical Field

[0001] This invention relates to the field of elevator safety control technology, and more specifically, to a remote functional safety device and control method for escalators based on a safety PLC. Background Technology

[0002] With the improvement of people's living standards, escalators and moving walkways are widely used in shopping malls, subway stations, airports and other places, providing convenience for people's lives. However, many safety-related accidents have also occurred, such as elderly people and children falling while riding escalators or moving walkways, people collapsing due to crowds, steps sinking and trapping people, steps overturning, and floor slabs sinking. Although various manufacturers have taken various safety protection measures, because escalators and moving walkways are long, it is difficult to monitor every part in an all-round way. As a result, when accidents occur, escalators or moving walkways fail to stop in time, and safety accidents causing injuries and deaths occur from time to time.

[0003] Escalators, as core transportation equipment in public transportation and commercial venues, are directly related to passenger safety. Existing escalator safety technologies have significant shortcomings:

[0004] When an accident occurs, on-site personnel must trigger the emergency stop button next to the steps. There is no effective remote control method. If there are no personnel on-site, the response delay may easily expand the scope of casualties. Existing remote control devices mostly use a combination of ordinary PLCs and relays, which do not meet the SIL2 level requirements of PESSRAE (Programmable Electronic Safety Related Systems), lack redundancy design and self-diagnostic functions, and pose a risk of malfunction or failure. Most escalators use independent control, which cannot achieve centralized monitoring and unified scheduling of multiple escalators, which is not conducive to the safety management of large venues.

[0005] To address the aforementioned issues, there is an urgent need for a remote safety control solution for escalators that combines AI-powered intelligent monitoring, reliable long-distance transmission, high security levels, and centralized management functions, in order to resolve the problems existing in the current technologies. Summary of the Invention

[0006] The present invention aims to solve the problem of how to remotely and safely control escalators.

[0007] To address the aforementioned issues, this invention provides a remote functional safety device for escalators based on a safety PLC, comprising: an AI abnormal behavior detection module for real-time acquisition of escalator area images, identification of dangerous behaviors such as passenger falls, crowding, and climbing, and triggering alarms to provide "abnormal evidence" for subsequent remote control; a main control room monitoring and output module for remote decision-making after confirming an anomaly and outputting a remotely transmittable AI enable signal, while simultaneously alerting the operator with audible and visual alarms to prevent misoperation; a signal transmission module for converting the AI ​​enable signal from the main control room from an electrical signal to a fiber optic signal for long-distance transmission, and then back to an electrical signal for transmission to the escalator end, solving the problems of low-voltage signal transmission attenuation and poor anti-interference; an escalator control cabinet detection module for converting the DC110V high-voltage AI enable signal from the main control room into a recognizable DC24V logic signal to ensure that the escalator end can accurately receive remote control commands; and a safety circuit control module for cutting off the escalator safety circuit to cause an emergency stop of the escalator, while ensuring reliable stopping action and preventing malfunctions from causing stopping failures.

[0008] The present invention provides a remote functional safety device for escalators based on a safety PLC, which, compared with the prior art, has the following beneficial effects, but is not limited to:

[0009] The various modules in the PLC-based remote functional safety device for escalators can work collaboratively and in close coordination. The AI ​​abnormal behavior detection module, acting as the "sensing end," can capture images and identify dangerous behaviors such as passenger falls, transmitting abnormal signals to the PC server and triggering alarms, providing a basis for remote control. The main control room monitoring and output module, acting as the "decision and command end," is used to make remote decisions after confirming abnormalities and output AI enable signals that can be transmitted over long distances. Simultaneously, it alerts the operator through audible and visual alarms to prevent misoperation. The signal transmission module, acting as the "transmission channel," is used to transmit the AI ​​enable signals issued by the main control room. The system converts electrical signals into fiber optic signals for long-distance transmission, and then converts them back into electrical signals for transmission to the escalator end, solving the problems of low-voltage signal attenuation and poor anti-interference. The escalator control cabinet detection module acts as a "signal conversion terminal," converting the DC110V high-voltage AI enable signal from the main control room into a recognizable DC24V logic signal, ensuring that the escalator end can accurately receive remote control commands. The safety circuit control module acts as an "execution terminal," cutting off the escalator safety circuit to stop the escalator in an emergency, while ensuring reliable stopping action and preventing malfunctions that could lead to stopping failure, thereby achieving the goal of remotely controlling the escalator for emergency stopping.

[0010] Furthermore, the AI ​​abnormal behavior detection module includes a network camera, a network hard disk recorder, an AIBOX intelligent algorithm box, and a PC server. The network camera and the network hard disk recorder are connected via a network cable. The network hard disk recorder is connected to the AIBOX intelligent algorithm box and the PC server via network cables. The PC server is connected to the main control room monitoring and output module via Ethernet.

[0011] Furthermore, the main control room monitoring and output module includes an emergency stop switch, a main control room safety PLC, a dual-output safety relay, a DC110V output switching power supply, and a three-color audible and visual alarm indicator. The emergency stop switch, the dual-output safety relay, and the three-color audible and visual alarm indicator are electrically connected to the main control room safety PLC, and the DC110V output switching power supply is electrically connected to the dual-output safety relay.

[0012] Furthermore, the signal transmission module includes an Ethernet-to-fiber optic module, a fiber optic-to-Ethernet module, and a single-mode fiber optic cable. The Ethernet-to-fiber optic module is connected to the safety PLC in the main control room via a network cable, the fiber optic-to-Ethernet module is connected to the escalator control cabinet detection module via a network cable, and the Ethernet-to-fiber optic module and the fiber optic-to-Ethernet module are connected via a single-mode fiber optic cable.

[0013] Furthermore, the escalator control cabinet detection module includes a dual-input safety relay and an escalator control cabinet safety PLC, wherein the dual-input safety relay is electrically connected to the escalator control cabinet safety PLC.

[0014] Furthermore, the safety circuit control module includes a dual-channel series-output safety circuit control relay, which is electrically connected to the safety PLC of the escalator control cabinet, and the dual-channel series-output safety circuit control relay is connected in series to the escalator safety circuit.

[0015] Furthermore, the pedestrian images captured by the network camera are stored in the network hard disk video recorder. The network hard disk video recorder sends the image information to the AIBOX intelligent algorithm box for AI processing and analysis, and reads the analysis results back to storage. The escalator remote monitor software in the PC server displays the information.

[0016] Furthermore, the normally open contact of the emergency stop switch is connected to the T4 and SI1 ports of the main control room safety PLC, and the normally closed contact is connected to the T5 and SI0 ports of the main control room safety PLC; the first terminal of the dual-output safety relay coil is connected to the SO2 port of the main control room safety PLC, and the second terminal of the dual-output safety relay coil is connected to 24V ground; the normally closed contact of the dual-output safety relay is connected to the T0 and SI1 ports of the main control room safety PLC, the first terminal of the normally open contact is connected to DC110V, and the second terminal is connected to the AI ​​enable output signal for long-distance transmission.

[0017] Furthermore, one end of the coil of the dual-input safety relay is connected to the high-voltage AI enable signal transmitted over a long distance, and the other end of the coil is connected to DC110V ground. The two normally closed contacts are connected to T1, SI0, and SI8 of the escalator control cabinet safety PLC, and the two normally open contacts are connected to T2, SI1, and SI9 of the escalator control cabinet safety PLC. One end of the coil of the dual-series output safety circuit control relay is connected to SO1 and SO3 of the escalator control cabinet safety PLC, and the other end of the coil is connected to DC24V ground. The two normally closed feedback contacts are connected to T3, SI2, and SI5 of the escalator control cabinet safety PLC, and the two normally open contacts are connected in series and then fed into the escalator safety circuit.

[0018] A control method for a remote functional safety device of an escalator using a safety PLC includes the following steps:

[0019] Step 1: The network camera captures images of passenger behavior on the escalator in real time and transmits them to the network hard disk recorder, which then pushes them to the AIBOX smart algorithm box. The AIBOX smart algorithm box performs anomaly recognition on the images. If it detects passenger falling, crowding, or climbing behavior, it generates an anomaly signal and transmits it to the PC server, triggering an audible and visual alarm.

[0020] Step 2: After the operator in the monitoring room confirms the abnormality through the video screen of the PC server, he presses the emergency stop switch; the safety PLC in the main control room simultaneously receives the AI ​​abnormality signal and the emergency stop switch signal, determines that the dual signals are enabled, and outputs DC24V voltage to the dual-output safety relay.

[0021] Step 3: The dual-output safety relay coil is energized, the normally open contact closes, and the voltage of the DC110V output switching power supply is output as an AI enable signal through the contact. The AI ​​enable signal is converted into an optical fiber signal by the Ethernet to fiber optic module and transmitted to the fiber optic to Ethernet module of the escalator control cabinet through a single-mode optical fiber.

[0022] Step 4: The fiber optic to Ethernet module converts the fiber optic signal back into an electrical signal, energizes the dual-input safety relay coil, converts the DC110V signal into a DC24V logic signal, and transmits it to the safety PLC in the escalator control cabinet.

[0023] Step 5: The safety PLC in the escalator control cabinet detects a valid signal, cuts off the coil voltage of the dual-channel series output safety circuit control relay, the normally open contact of the relay opens, cuts off the escalator safety circuit, and the escalator stops running in an emergency.

[0024] Step 6: The safety PLC in the escalator control cabinet detects the escalator stop status through feedback contacts and transmits the "stopped" feedback signal back to the safety PLC in the main control room via optical fiber; the PC server displays the escalator stop status and automatically stores fault information. Attached Figure Description

[0025] Figure 1 This is a system module diagram of a remote functional safety device for escalators based on a safety PLC, according to an embodiment of the present invention.

[0026] Figure 2 This is a system block diagram of a remote functional safety device for escalators based on a safety PLC, according to an embodiment of the present invention.

[0027] Figure 3 for Figure 2 A magnified structural diagram of area A in the diagram;

[0028] Figure 4 for Figure 2 A magnified structural diagram of region B in the diagram;

[0029] Figure 5 for Figure 2 A magnified structural diagram of region C in the diagram;

[0030] Figure 6 This is a control logic diagram of the PLC relay output of a remote functional safety device for escalators based on a safety PLC, according to an embodiment of the present invention.

[0031] Figure 7 This is a flowchart of the PLC relay output control program for a remote functional safety device for escalators based on a safety PLC, according to an embodiment of the present invention.

[0032] Figure 8 This is a PLC relay output control logic diagram of an escalator control cabinet based on a safety PLC, according to an embodiment of the present invention.

[0033] Figure 9 This is a flowchart of the PLC relay output control program for an escalator control cabinet based on a safety PLC, according to an embodiment of the present invention.

[0034] Figure 10 This is a schematic diagram of the main control room circuit of a remote functional safety device for escalators based on a safety PLC, according to an embodiment of the present invention.

[0035] Figure 11 This is a circuit diagram of an escalator control cabinet for a remote functional safety device for escalators based on a safety PLC, according to an embodiment of the present invention.

[0036] Figure 12 This is a flowchart illustrating a control method for a remote functional safety device of an escalator using a safety PLC, according to an embodiment of the present invention.

[0037] Explanation of reference numerals in the attached figures:

[0038] M1, AI Abnormal Behavior Detection Module; M2, Main Control Room Monitoring and Output Module; M3, Signal Transmission Module; M4, Escalator Control Cabinet Detection Module; M5, Safety Circuit Control Module; 1, Network Camera; 2, Network Hard Disk Recorder; 3, AIBOX Intelligent Algorithm Box; 4, Main Control Room Safety PLC; 5, Dual-Output Safety Relay; 6, Ethernet to Fiber Optic Module; 7, Dual-Input Safety Relay; 8, Fiber Optic to Ethernet Module; 9, Escalator Control Cabinet Safety PLC; 10, Dual-Series Output Safety Circuit Control Relay; 12, DC110V Output Switching Power Supply; 15, PC Server; 16, Emergency Stop Switch. Detailed Implementation

[0039] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings showing multiple embodiments according to this application. It should be understood that the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. All other embodiments obtained by those skilled in the art based on the embodiments described in this application without creative effort will fall within the scope of protection of this application.

[0040] Unless otherwise defined, all technical and scientific terms used in this application have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used in the description of this application is for the purpose of describing specific embodiments only and is not intended to limit this application; the terms "comprising," "including," "having," "containing," etc., in the description, claims, and accompanying drawings of this application are open-ended terms. Therefore, "comprising," "including," or "having" refers to, for example, a method or apparatus having one or more steps or elements, but is not limited to having only these one or more elements. The terms "first," "second," etc., in the description, claims, or accompanying drawings of this application are used to distinguish different objects, not to describe a specific order or hierarchy. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0041] It should be emphasized that when the term "comprising / including" is used in this specification, it is used to explicitly indicate the presence of the stated feature, integer, step, or component, but does not exclude the presence or addition of one or more other features, integers, steps, parts, or groups of features, integers, steps, or parts.

[0042] See Figures 1-6 This invention discloses a remote functional safety device for escalators based on a safety PLC, comprising: an AI abnormal behavior detection module M1, used to collect real-time images of the escalator area, identify dangerous behaviors such as passenger falls, crowding, and climbing, and trigger alarms to provide "abnormal evidence" for subsequent remote control; a main control room monitoring and output module M2, used to make remote decisions after confirming an anomaly, and output an AI enable signal that can be transmitted over long distances, while simultaneously alerting the operator through audible and visual alarms to avoid misoperation; a signal transmission module M3, used to convert the AI ​​enable signal sent from the main control room from an electrical signal to an optical fiber signal for long-distance transmission, and then convert it back to an electrical signal to be transmitted to the escalator end, solving the problems of low-voltage signal transmission attenuation and poor anti-interference; an escalator control cabinet detection module M4, used to convert the DC110V high-voltage AI enable signal from the main control room into a recognizable DC24V logic signal to ensure that the escalator end can accurately receive remote control commands; and a safety circuit control module M5, used to cut off the escalator safety circuit, causing the escalator to stop urgently, while ensuring reliable escalator stopping action and avoiding malfunctions that could lead to escalator stopping failure.

[0043] In this embodiment, the modules can collaborate and work closely together. The AI ​​abnormal behavior detection module, acting as the "sensing end," can capture images and identify dangerous behaviors such as passengers falling, transmitting abnormal signals to the PC server and triggering alarms to provide a basis for remote control. The main control room monitoring and output module, acting as the "decision and command end," is used to make remote decisions after confirming abnormalities and output AI enable signals that can be transmitted over long distances. It also alerts the operator through audible and visual alarms to prevent misoperation. The signal transmission module, acting as the "transmission channel," converts the AI ​​enable signals issued by the main control room from electrical signals into electronic signals. Fiber optic signals are transmitted over long distances and then converted back into electrical signals for transmission to the escalator end, solving the problems of low-voltage signal attenuation and poor anti-interference. The escalator control cabinet detection module acts as a "signal conversion terminal," converting the DC110V high-voltage AI enable signal from the main control room into a recognizable DC24V logic signal, ensuring that the escalator end can accurately receive remote control commands. The safety circuit control module acts as an "execution terminal," cutting off the escalator safety circuit to stop the escalator in an emergency, while ensuring reliable stopping action and preventing malfunctions that could lead to stopping failure, thereby achieving the goal of remotely controlling the escalator for emergency stopping.

[0044] Optionally, the AI ​​abnormal behavior detection module M1 includes a network camera 1, a network hard disk recorder 2, an AIBOX intelligent algorithm box 3, and a PC server 15. The network camera 1 and the network hard disk recorder 2 are connected via a network cable. The network hard disk recorder 2 is connected to the AIBOX intelligent algorithm box 3 and the PC server 15 via network cables. The PC server 15 is connected to the main control room monitoring and output module M2 via an Ethernet.

[0045] In this embodiment, the network camera 1 captures dynamic images through a CMOS sensor, converts the analog signal into a digital stream, and transmits it to the network hard disk recorder 2 via a network cable; the network hard disk recorder 2 simultaneously pushes the stream to the AIBOX intelligent algorithm box 3, which identifies anomalies through "human posture key point extraction (such as head, torso, and limb coordinates) + density clustering analysis"; after identifying anomalies, the algorithm box outputs a 5V high-level signal to the PC server 15 to trigger an alarm.

[0046] Among them, network cameras 1 are evenly distributed along the length of the escalator (5-8m spacing), supporting 1080P (1920×1080) high-definition shooting with a frame rate ≥30fps. They possess wide dynamic range capabilities, adapting to complex lighting environments such as strong light and shadow, and capture real-time images of escalator passenger behavior, covering the entire area including steps, handrails, and platforms. A network hard disk recorder 2 connects to network cameras 1 via a Cat5e cable, supporting 8-16 video inputs with a storage capacity ≥4TB. It features image browsing, recording storage, playback, and alarm linkage functions, synchronously transmitting the captured video stream to the AIBOX intelligent algorithm box 3 and the PC server 15. The AIBOX intelligent algorithm... The AIBOX intelligent algorithm box 3 is equipped with an NVIDIA Jetson AGXXavier processor and has a built-in pedestrian behavior recognition algorithm optimized based on YOLOv5. It can simultaneously connect to 8 monitoring front-ends to identify abnormal behaviors such as passengers falling, crowding, and climbing handrails. After abnormal identification, it generates a high-level 5V abnormal signal. The PC server 15 is equipped with the Windows Server 2019 operating system and escalator remote monitor software. It is connected to the network hard disk recorder 2 via a gigabit network cable. It can display video in real time, receive abnormal signals from the AIBOX intelligent algorithm box 3 and pop up a red alarm information box. At the same time, it transmits the abnormal signals to the safety PLC 4 in the main control room via Ethernet.

[0047] The specific optimizations to the built-in pedestrian behavior recognition algorithm based on YOLOv5 are as follows: An attention mechanism, namely the CBAM attention module, is introduced into the C3 module of YOLOv5s. Specifically, it is located after the 6th layer of the C3 module. The channel attention weight parameter of the newly added CBAM module is set to 0.8, and the spatial attention convolution kernel size is 3×3. By strengthening the extraction of key behavioral features, the recognition ability of non-standard postures such as falls and climbing is improved. At the same time, for pedestrian targets in escalator scenarios with relatively large scales, the algorithm is optimized. To maintain consistent characteristics, the original small-scale detector head (13×13) of YOLOv5s was removed, while the medium-scale (26×26) and large-scale (52×52) detector heads were retained. The anchor frame sizes were adjusted to [10,13], [16,30], [33,23], [30,61], [62,45], and [59,119] to accommodate the height and behavioral posture range of pedestrians within the escalator area. For training parameter settings, a public dataset (a subset of the COCO dataset for pedestrians) and a self-built escalator field were used. The dataset was constructed by combining scene and dataset data. The self-built dataset contains 8,000 images of pedestrian behavior on escalators in different locations such as shopping malls, subway stations, and airports. Among them, there are 5,000 images of normal behavior and 3,000 images of abnormal behavior such as falling, crowding, and climbing. All images are labeled with bounding boxes and behavior category labels in YOLO format. For training hyperparameters, the batch size was set to 16, the initial learning rate was 0.001, and a cosine annealing learning rate decay strategy was adopted with a decay period of 10 epochs and a weight decay coefficient of 0.0005. The number of training iterations was 100 epochs, with the first 5 epochs serving as warm-up training, where the warm-up learning rate linearly increased from 0.0001 to 0.001. The loss function was adjusted, using the CIoU loss function to calculate the bounding box regression loss and FocalLoss for classification loss. The FocalLoss focusing parameter γ was set to 2, and the balance factor α was set to 0.25 to address the class imbalance problem caused by the small number of abnormal behavior samples. In the model inference optimization stage, TensorRT is used to quantize and accelerate the trained model with a quantization precision of INT8. The confidence threshold during inference is initially set to 0.5, and the IOU threshold for non-maximum suppression is set to 0.45 to ensure that the inference speed on the AIBOX intelligent algorithm box is ≥30fps, which meets the real-time detection requirements.

[0048] Meanwhile, differentiated confidence thresholds were set for different abnormal behavior types. Because the posture characteristics of falling behavior are obvious, the threshold is higher to ensure recognition accuracy. The confidence threshold for "falling" behavior is set at 0.65. In crowded scenarios, there is a lot of overlap among pedestrian targets, so the threshold is appropriately lowered to avoid missed detections. The confidence threshold for "crowding" behavior is set at 0.55. Climbing behavior involves a large range of motion, so it is necessary to balance accuracy and the false negative rate. The confidence threshold for "climbing" behavior is set at 0.60. At the same time, the algorithm supports dynamic adjustment of thresholds based on the real-time pedestrian density of escalators. When the pedestrian density is ≥5 people / ㎡, the threshold for crowding behavior is automatically lowered to 0.50, while the thresholds for other behaviors remain unchanged. When the pedestrian density is <1 person / ㎡, the thresholds for all behaviors are raised by 0.05 to reduce false alarms in single-person scenarios. The false alarm filtering rules are set as follows: First, there is time continuity filtering. After detecting suspected abnormal behavior, it is necessary to meet the confidence threshold of the corresponding behavior for 3 consecutive frames before it is judged as a valid abnormal signal, so as to avoid false alarms caused by blurry single-frame images or misjudgment of pedestrian instantaneous posture. Second, there is spatial rationality filtering. An effective recognition area is set based on the physical structure of the escalator. An alarm is triggered only when the center point of the detected abnormal behavior target is located in the effective area. In addition, there is behavioral feature verification. For the "falling" behavior, the angle relationship of the target's posture key points (head, torso, limbs) is further verified. A fall is confirmed only when the angle between the torso and the ground is ≤30° and lasts for more than 2 frames. For the "climbing" behavior, it is verified whether the target is in the escalator handrail area (preset handrail coordinate range) and the vertical displacement is ≥0.5m / second, so as to avoid misjudging normal actions such as raising hands or bending over as climbing.

[0049] Optionally, the main control room monitoring and output module M2 includes an emergency stop switch 16, a main control room safety PLC 4, a dual-output safety relay 5, a DC110V output switching power supply 12, and a three-color audible and visual alarm indicator. The emergency stop switch 16, the dual-output safety relay 5, and the three-color audible and visual alarm indicator are electrically connected to the main control room safety PLC 4, and the DC110V output switching power supply 12 is electrically connected to the dual-output safety relay 5.

[0050] In this embodiment, the main control room monitoring and output module M2 includes: a Schneider XB2-BS542C emergency stop switch (dual contact), 6 Omron CC20N safety PLCs, 48 ​​Omron G7SA-5A1B dual-output safety relays, a Mean Well LRS-150-110DC110V switching power supply, and an LTE-5101 three-color audible and visual alarm light. The emergency stop switch has a 1.5mm... 2 The copper core wire is connected to the PLC input port, the relay coil is connected to the PLC output port, and the power output terminal is 2.5mm thick. 2 Connect the wire to the normally open contact of the relay.

[0051] In this system, the emergency stop switch 16 normally has its normally open contact open and normally closed contact closed, and the PLC detects a "low level + low level" signal. When it receives an AI abnormal signal (high level) and the operator presses the emergency stop switch (changing to "high level + high level"), the PLC determines that both signals are valid and outputs 24V to the relay coil. After the relay operates, the normally closed contact opens to provide feedback, and the normally open contact closes to output a 110V enable signal, while simultaneously triggering an audible and visual alarm (85dB / 2Hz flashing). This dual-signal confirmation mechanism completely avoids accidental escalator stops. A network of 6 PLCs enables zoned control of 24 escalators, with a single PLC response time of ≤0.3 seconds, resulting in a 6-fold increase in reliability compared to a single PLC solution.

[0052] Optionally, the signal transmission module M3 includes an Ethernet-to-fiber optic module 6, a fiber optic-to-Ethernet module 8, and a single-mode fiber optic cable. The Ethernet-to-fiber optic module 6 is connected to the main control room safety PLC 4 via a network cable. The fiber optic-to-Ethernet module 8 is connected to the escalator control cabinet detection module M4 via a network cable. The Ethernet-to-fiber optic module 6 and the fiber optic-to-Ethernet module 8 are connected via a single-mode fiber optic cable.

[0053] In this embodiment, the signal transmission module M3 uses the H3C HF-450S Ethernet to fiber optic module, combined with G652D single-mode fiber (2 cores / 9μm fiber core). The main control room module is connected to the PLC Ethernet port via a gigabit network cable, and the escalator end module is connected to the escalator control cabinet PLC. The fiber is laid along the cable tray (bending radius ≥30mm), and both ends are spliced ​​using SC connectors.

[0054] The 110V electrical signal in the main control room is converted into a 1310nm wavelength optical signal via an Ethernet-to-fiber optic module and transmitted with low loss (0.5dB / km) through single-mode fiber. The escalator end module converts the optical signal back into an electrical signal, achieving attenuation-free transmission over a distance of 5km and avoiding signal loss caused by electromagnetic interference.

[0055] Optionally, the escalator control cabinet detection module M4 includes a dual-input safety relay 7 and an escalator control cabinet safety PLC 9, wherein the dual-input safety relay 7 is electrically connected to the escalator control cabinet safety PLC 9.

[0056] In this embodiment, the escalator control cabinet detection module M4 includes an Omron G7SA-5A1B dual-input safety relay and an Omron CC20N safety PLC. One end of the relay coil is connected to the 110V signal output from the fiber optic module, and the other end is connected to 110V ground; its normally closed contacts are connected to the PLC's T1 / SI0 and T1 / SI8 ports, and its normally open contacts are connected to the T2 / SI1 and T2 / SI9 ports. The wires used are 1mm². 2 Copper core wire.

[0057] The 110V enable signal drives the relay coil to be energized, the normally closed contact opens (PLC detects a high level) and the normally open contact closes (PLC detects a low level). The high voltage signal is converted into a 24V logic signal that the PLC can recognize through the level change. The PLC confirms the validity of the signal by comparing the status of the two contacts (normally closed open + normally open closed). An alarm is triggered when a single contact fails.

[0058] Optionally, the safety circuit control module M5 includes a dual-channel series output safety circuit control relay 10, which is electrically connected to the safety PLC9 of the escalator control cabinet, and the dual-channel series output safety circuit control relay 10 is connected in series to the escalator safety circuit.

[0059] In this embodiment, the safety circuit control module M5 uses two Omron G7SA-5A1B safety relays to form a dual-channel series output. The coils are connected to the SO1 and SO3 ports of the escalator control cabinet PLC, respectively, and the other end is connected to 24V ground. The normally open contacts are connected in series to the escalator safety circuit (between the emergency stop circuit and the safety contactor), and the normally closed feedback contacts are connected to the T3 / SI2 and T3 / SI5 ports of the PLC.

[0060] During normal operation, the PLC outputs 24V to energize the relay coil, and the normally open contact closes to keep the safety circuit open. After receiving the stop command, the PLC cuts off the output, the coil loses power, causing the normally open contact to open, the safety circuit is cut off, and the escalator is braked. The PLC monitors the contact status through feedback contacts, and immediately alarms when any relay fails.

[0061] Optionally, the pedestrian images captured by the network camera 1 are stored in the network hard disk video recorder 2. The network hard disk video recorder 2 sends the image information to the AIBOX intelligent algorithm box 3 for AI processing and analysis, and reads the analysis results back to storage. The escalator remote monitor software in the PC server 15 displays the information.

[0062] In this embodiment, the 1080P images captured by the network camera are stored in H.265 format in the network video recorder (with a 30-day loop coverage period); the video recorder pushes the real-time bitstream to AIBOX via the RTSP protocol, and the abnormal results (including timestamps and behavior types) generated after processing by the algorithm box are sent back to the video recorder for storage via the SDK; the escalator remote monitor software (developed in C#) installed on the PC reads the video recorder data through the API and displays the images of 24 escalators and abnormal indicators in a split-screen interface.

[0063] Image storage adopts a dual backup mechanism of "raw video + analysis results". The recorder associates video clips with abnormal events through indexes. The software refreshes the device status periodically (at 1-second intervals). When an abnormality occurs, it automatically jumps to the corresponding screen and pops up a prompt. It supports searching historical data by time / type.

[0064] Optionally, the normally open contact of the emergency stop switch 16 is connected to the T4 and SI1 ports of the main control room safety PLC4, and the normally closed contact is connected to the T5 and SI0 ports of the main control room safety PLC4; the first end of the coil of the dual-output safety relay 5 is connected to the SO2 port of the main control room safety PLC4, and the second end of the coil of the dual-output safety relay 5 is connected to 24V ground; the normally closed contact of the dual-output safety relay 5 is connected to the T0 and SI1 ports of the main control room safety PLC4, the first end of the normally open contact is connected to DC110V, and the second end is connected to the AI ​​enable output signal for long-distance transmission.

[0065] In this embodiment, the normally open contact (NO) of the emergency stop switch 16 is 1.5mm wide. 2 The wires connect T4 (common terminal) and SI1 (input) of the PLC in the main control room, and the normally closed contact (NC) connects T5 and SI0; one end of the coil of the dual-output safety relay (5) is connected to SO2 (output) of the PLC, and the other end is connected to 24V ground (0V); the normally closed contact of the relay is connected to T0 and SI1 of the PLC, one end of the normally open contact is connected to the positive terminal of 110V power supply, and the other end is used as the AI ​​enable signal output terminal (connected to the fiber optic module).

[0066] The PLC determines the status by monitoring the SI0 / SI1 port levels: under normal conditions, SI0=0V (NC closed) and SI1=0V (NO open); after an emergency stop is triggered, SI0=24V (NC open) and SI1=24V (NO closed). Combined with the AI ​​abnormal signal (SO1=24V), the PLC drives SO2 to output 24V to activate the relay. At the same time, it monitors the open status of the normally closed contact through the SI1 port (confirming the action). The hardware wiring achieves "dual signal interlocking" to avoid malfunctions caused by software misjudgment. The contact status feedback reduces the fault location time to 5 minutes and improves maintenance efficiency by 80%.

[0067] Optionally, one end of the coil of the dual-input safety relay 7 is connected to a high-voltage AI enable signal transmitted over a long distance, and the other end of the coil is connected to DC110V ground. The two normally closed contacts are connected to T1, SI0, and SI8 of the escalator control cabinet safety PLC9, and the two normally open contacts are connected to T2, SI1, and SI9 of the escalator control cabinet safety PLC9. One end of the coil of the dual-series output safety circuit control relay 10 is connected to SO1 and SO3 of the escalator control cabinet safety PLC9, and the other end of the coil is connected to DC24V ground. The two normally closed feedback contacts are connected to T3, SI2, and SI5 of the escalator control cabinet safety PLC9, and the two normally open contacts are connected in series and then inserted into the escalator safety circuit.

[0068] In this embodiment, one end of the coil of the dual-input safety relay 7 is connected to a 110V enable signal, and the other end is connected to 110V ground; its first set of normally closed contacts is connected to T1 and SI0 of the escalator PLC, and its second set of normally closed contacts is connected to T1 and SI8; its first set of normally open contacts is connected to T2 and SI1, and its second set of normally open contacts is connected to T2 and SI9. The coil of the dual-output safety relay 10 is connected to SO1, SO3 and 24V ground respectively; its feedback normally closed contacts are connected to T3 and SI2, and T3 and SI5; the normally open contacts are connected in series and then connected to the safety circuit.

[0069] After the 110V signal causes the relay to operate, SI0 / SI8 becomes high (normally closed open), and SI1 / SI9 becomes low (normally open closed). The PLC determines the signal to be valid through "double normally closed open + double normally open closed". When the elevator stops, the PLC cuts off the output of SO1 / SO3, the normally open contact of the relay opens to cut off the safety circuit, and at the same time SI2 / SI5 monitors the contact status to ensure reliable operation.

[0070] Please see Figure 7 The control method of a remote functional safety device for escalators using a safety PLC, provided in this embodiment of the invention, is described in detail below:

[0071] S100: The network camera 1 collects images of passenger behavior on the escalator in real time and transmits them to the network hard disk recorder 2, which then pushes them to the AIBOX smart algorithm box 3. The AIBOX smart algorithm box 3 performs anomaly recognition on the images. If it detects passenger falling, crowding, or climbing behavior, it generates an abnormal signal and transmits it to the PC server 15, triggering an audible and visual alarm.

[0072] S200: After the operator in the monitoring room confirms the abnormality through the video screen of the PC server 15, he presses the emergency stop switch 16; the main control room safety PLC4 simultaneously receives the AI ​​abnormality signal and the emergency stop switch signal, determines that the dual signal is enabled, and outputs DC24V voltage to the dual-output safety relay 5.

[0073] S300: When the coil of the dual-output safety relay 5 is energized, the normally open contact closes, and the voltage of the DC110V output switching power supply 12 is output through the contact to output the AI ​​enable signal; the AI ​​enable signal is converted into an optical fiber signal by the Ethernet to fiber optic module 6, and transmitted to the fiber optic to Ethernet module 8 of the escalator control cabinet through a single-mode optical fiber.

[0074] S400: The fiber optic to Ethernet module 8 restores the fiber optic signal to an electrical signal. The coil of the dual-input safety relay 7 is energized, converting the DC110V signal into a DC24V logic signal and transmitting it to the safety PLC 9 of the escalator control cabinet.

[0075] S500: When the safety PLC9 of the escalator control cabinet detects a valid signal, it cuts off the coil voltage of the dual-channel series output safety circuit control relay 10, causing the normally open contact of the relay to open, cutting off the escalator safety circuit, and the escalator to stop running in an emergency.

[0076] S600: The safety PLC9 in the escalator control cabinet detects the escalator stop status through feedback contacts and transmits the "stopped" feedback signal back to the safety PLC4 in the main control room via optical fiber; the PC server 15 displays the escalator stop status and automatically stores fault information including the abnormal time, escalator number, and stop time.

[0077] Take the scenario of a passenger falling on a subway escalator as an example:

[0078] First, the central camera captures the fall, and the video stream is simultaneously transmitted to the AIBOX via a recorder. The algorithm box identifies the anomaly within 1.2 seconds and sends a signal to the PC, triggering an audible and visual alarm. After operator confirmation, the emergency stop switch is pressed, and the PLC in the main control room outputs 24V to the relay within 0.3 seconds. The relay activates and outputs a 110V signal, which, after conversion by the fiber optic module, takes ≤20ms to transmit over 5km. The relay at the escalator end converts the signal into a 24V logic signal, and the PLC confirms its validity within 0.2 seconds. The dual-output relays disconnect, the safety circuit is cut off, and the escalator stops within 1.2 seconds. The stopped status is transmitted back to the main control room via fiber optic cable, and the PC displays the status and stores the fault record (including 0.5x playback clips).

[0079] Safety control is achieved through "layered processing + closed-loop feedback": the AI ​​abnormal behavior detection module M1 completes abnormal identification, the main control room monitoring and output module M2 achieves dual confirmation, the signal transmission module M3 ensures the reliability of commands, the escalator control cabinet detection module M4 adapts to signal types, and the safety loop control module M5 completes the escalator stop action. The status of each link is fed back in real time to form a closed loop; the whole process takes ≤3 seconds, which is 70% faster than the traditional "on-site discovery - manual emergency stop" mode (average 10 seconds); fault information is automatically archived for 90 days to meet the requirements for accident traceability, and the overall safety protection level reaches SIL2.

[0080] While the present invention has been disclosed above, its scope of protection is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention, and all such changes and modifications will fall within the scope of protection of the present invention.

Claims

1. A remote functional safety device for escalators based on a safety PLC, characterized in that, include: The AI ​​abnormal behavior detection module (M1) is used to collect images of the escalator area in real time, identify dangerous behaviors such as passengers falling, crowding, and climbing, and trigger alarms to provide "abnormal evidence" for subsequent remote control. The AI ​​abnormal behavior detection module (M1) includes a network camera (1), a network hard disk recorder (2), an AIBOX intelligent algorithm box (3), and a PC server (15). The network camera (1) and the network hard disk recorder (2) are connected by a network cable. The network hard disk recorder (2) is connected to the AIBOX intelligent algorithm box (3) and the PC server (15) by network cables. The PC server (15) is connected to the main control room monitoring and output module (M2) via Ethernet. The main control room monitoring and output module (M2) is used to make remote decisions after confirming an anomaly and output an AI enable signal that can be transmitted over a long distance. At the same time, it reminds the operator through sound and light alarms to avoid misoperation. The main control room monitoring and output module (M2) includes an emergency stop switch (16), a main control room safety PLC (4), a dual-output safety relay (5), a DC110V output switching power supply (12), and a three-color sound and light alarm indicator. The emergency stop switch (16), the dual-output safety relay (5), and the three-color sound and light alarm indicator are electrically connected to the main control room safety PLC (4), and the DC110V output switching power supply (12) is electrically connected to the dual-output safety relay (5). The signal transmission module (M3) is used to convert the AI ​​enable signal issued by the main control room from an electrical signal to an optical fiber signal for long-distance transmission, and then convert it back to an electrical signal to be transmitted to the escalator end, thus solving the problems of low voltage signal transmission attenuation and poor anti-interference. The signal transmission module (M3) includes an Ethernet to optical fiber module (6), an optical fiber to Ethernet module (8) and a single-mode optical fiber. The Ethernet to optical fiber module (6) is connected to the safety PLC (4) in the main control room via a network cable. The optical fiber to Ethernet module (8) is connected to the escalator control cabinet detection module (M4) via a network cable. The Ethernet to optical fiber module (6) and the optical fiber to Ethernet module (8) are connected via a single-mode optical fiber. The escalator control cabinet detection module (M4) is used to convert the DC110V high-voltage AI enable signal from the main control room into a recognizable DC24V logic signal to ensure that the escalator end can accurately receive remote control commands. The escalator control cabinet detection module (M4) includes a dual-input safety relay (7) and an escalator control cabinet safety PLC (9). The dual-input safety relay (7) is electrically connected to the escalator control cabinet safety PLC (9). The safety circuit control module (M5) is used to cut off the escalator safety circuit, so that the escalator stops in an emergency, while ensuring that the escalator stops reliably and avoiding failure to stop the escalator due to a malfunction. The safety circuit control module (M5) includes a dual-channel series output safety circuit control relay (10). The dual-channel series output safety circuit control relay (10) is electrically connected to the safety PLC (9) of the escalator control cabinet, and the dual-channel series output safety circuit control relay (10) is connected in series to the escalator safety circuit.

2. The remote functional safety device for escalators based on a safety PLC according to claim 1, characterized in that, The pedestrian images captured by the network camera (1) are stored in the network hard disk video recorder (2). The network hard disk video recorder (2) sends the image information to the AIBOX intelligent algorithm box (3) for AI processing and analysis, and reads the analysis results back to storage. The escalator remote monitor software in the PC server (15) displays the information.

3. The remote functional safety device for escalators based on a safety PLC according to claim 2, characterized in that, The normally open contact of the emergency stop switch (16) is connected to the T4 and SI1 ports of the main control room safety PLC (4), and the normally closed contact is connected to the T5 and SI0 ports of the main control room safety PLC (4); the first end of the coil of the dual-output safety relay (5) is connected to the SO2 port of the main control room safety PLC (4), and the second end of the coil of the dual-output safety relay (5) is connected to 24V ground; the normally closed contact of the dual-output safety relay (5) is connected to the T0 and SI1 ports of the main control room safety PLC (4), the first end of the normally open contact is connected to DC110V, and the second end is connected to the AI ​​enable output signal for long-distance transmission.

4. The remote functional safety device for escalators based on a safety PLC according to claim 3, characterized in that, One end of the coil of the dual-input safety relay (7) is connected to the high-voltage AI enable signal transmitted over a long distance, and the other end of the coil is connected to the DC110V ground. The two normally closed contacts are connected to T1, SI0 and SI8 of the escalator control cabinet safety PLC (9), and the two normally open contacts are connected to T2, SI1 and SI9 of the escalator control cabinet safety PLC (9). One end of the coil of the dual-series output safety circuit control relay (10) is connected to SO1 and SO3 of the escalator control cabinet safety PLC (9), and the other end of the coil is connected to the DC24V ground. The two normally closed feedback contacts are connected to T3, SI2 and SI5 of the escalator control cabinet safety PLC (9), and the two normally open contacts are connected in series and then connected to the escalator safety circuit.

5. A control method for a remote functional safety device for escalators using a safety PLC, based on the device described in any one of claims 1-4, characterized in that, Includes the following steps: Step 1: The network camera (1) collects images of escalator passenger behavior in real time and transmits them to the network hard disk recorder (2), and pushes them to the AIBOX intelligent algorithm box (3) simultaneously; the AIBOX intelligent algorithm box (3) performs abnormal identification on the images. If it detects passengers falling, crowding or climbing, it generates an abnormal signal and transmits it to the PC server (15) to trigger an audible and visual alarm. Step 2: After the operator in the monitoring room confirms the abnormality through the video screen of the PC server (15), he presses the emergency stop switch (16); the safety PLC (4) in the main control room receives the AI ​​abnormality signal and the emergency stop switch signal at the same time, determines that the dual signal is enabled, and outputs DC24V voltage to the dual-output safety relay (5). Step 3: The coil of the dual-output safety relay (5) is energized, the normally open contact is closed, and the voltage of the DC110V output switching power supply (12) is output through the contact to generate an AI enable signal; the AI ​​enable signal is converted into an optical fiber signal by the Ethernet to fiber optic module (6) and transmitted to the fiber optic to Ethernet module (8) of the escalator control cabinet through a single-mode optical fiber. Step 4: The fiber optic to Ethernet module (8) restores the fiber optic signal to an electrical signal, and the coil of the dual-input safety relay (7) is energized, converting the DC110V signal into a DC24V logic signal and transmitting it to the safety PLC (9) of the escalator control cabinet. Step 5: The safety PLC (9) of the escalator control cabinet detects a valid signal, cuts off the coil voltage of the dual-channel series output safety circuit control relay (10), the normally open contact of the relay opens, cuts off the escalator safety circuit, and the escalator stops running in an emergency. Step 6: The safety PLC (9) of the escalator control cabinet detects the escalator stop status through the feedback contact and transmits the "stopped" feedback signal back to the safety PLC (4) in the main control room via optical fiber; the PC server (15) displays the escalator stop status and automatically stores the fault information.

Images