Intelligent scheduling method and system of medical waste robot based on Internet of Things

Abstract

The invention discloses an intelligent scheduling method and system for a medical waste robot based on the Internet of Things, and relates to the technical field of the Internet of Things, and the method comprises the steps: installing an Internet of Things sensing equipment group on the medical waste robot; the robot is started, and standard medical waste multi-dimensional data flow is obtained through data processing; medical waste production prediction information is obtained, and meanwhile working state information of the robot is collected and obtained; determining a robot work scheduling scheme; and the robot is controlled to execute medical waste treatment and whole-process recording, a treatment data chain is obtained, and control is traced through the data chain. The technical problems that in the existing medical waste treatment process, manual operation efficiency is low, exposure risks exist, and waste treatment tasks are difficult to reasonably arrange according to actual conditions are solved, intelligent and automatic medical waste treatment is achieved, the treatment efficiency is improved, the occupational exposure risks of treatment personnel are reduced, and the medical waste treatment efficiency is improved. And reasonable scheduling can be realized according to the actual condition of the garbage, and the technical effect of whole-process fine management is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet of Things, and in particular to an intelligent scheduling method and system for a medical waste robot based on the Internet of Things. Background Art

[0002] Traditional medical waste disposal relies heavily on manual labor. Manual sorting, weighing, and transporting medical waste is not only inefficient and costly, but also exposes workers to direct contact with waste, exposing them to high occupational exposure risks.

[0003] While some simple auxiliary equipment currently exists for medical waste disposal, it lacks intelligent management and scheduling. Existing waste disposal methods struggle to accurately predict the volume, type, and peak hours of medical waste generation, making it difficult to rationally schedule treatment tasks based on actual conditions. Furthermore, traditional methods also lack the ability to record and trace the entire medical waste disposal process, resulting in a lack of transparency and management loopholes, making it difficult to meet the stringent requirements of the modern medical industry for efficient, safe, and environmentally friendly waste disposal. Summary of the Invention

[0004] This application solves the technical problems of low manual operation efficiency, exposure risks, and difficulty in rationally arranging waste disposal tasks according to actual conditions in the existing medical waste disposal process. This application installs an IoT sensing device group on a medical waste robot to collect multi-dimensional data streams of medical waste. Through data cleaning and waste production prediction, key information is obtained. Based on the working status of the medical waste robot, an ant colony algorithm is used to determine the scheduling plan, control the robot to perform tasks, and record the entire process to form a data chain for traceability and control. This achieves intelligent and efficient scheduling of medical waste disposal and refined management of the entire process, making the medical waste disposal process more scientific, safe, and environmentally friendly.

[0005] In response to the above technical problems, this application proposes a technical solution for an intelligent scheduling method and system of medical waste robots based on the Internet of Things.

[0006] In a first aspect, the application provides an intelligent scheduling method of medical waste robots based on Internet of Things, wherein the method comprises: installing a group of Internet of Things sensing devices on a medical waste robot, the group of Internet of Things sensing devices comprising a weight sensor, a type identification sensor, a GPS positioning device, an environmental sensor and a video monitoring device; starting the medical waste robot and collecting medical waste multi-dimensional data stream in real time through the group of Internet of Things sensing devices, performing data cleaning processing on the medical waste multi-dimensional data stream to obtain standard medical waste multi-dimensional data stream; performing waste production prediction on the standard medical waste multi-dimensional data stream to obtain medical waste production prediction information, while collecting working state information of the medical waste robot; performing scheduling scheme analysis based on the medical waste production prediction information and the working state information of the medical waste robot to determine a robot working scheduling scheme; controlling the medical waste robot to perform medical waste treatment and whole-process recording according to the robot working scheduling scheme to obtain a robot medical waste treatment data chain, and performing whole-process traceability management through the robot medical waste treatment data chain.

[0007] In a second aspect, the application provides an intelligent scheduling system of medical waste robots based on Internet of Things, wherein the system comprises: a sensing device installation module for installing a group of Internet of Things sensing devices on a medical waste robot, the group of Internet of Things sensing devices comprising a weight sensor, a type identification sensor, a GPS positioning device, an environmental sensor and a video monitoring device; a data stream obtaining module for starting the medical waste robot and collecting medical waste multi-dimensional data stream in real time through the group of Internet of Things sensing devices, performing data cleaning processing on the medical waste multi-dimensional data stream to obtain standard medical waste multi-dimensional data stream; a working information collecting module for performing waste production prediction on the standard medical waste multi-dimensional data stream to obtain medical waste production prediction information, while collecting working state information of the medical waste robot; a scheduling scheme determining module for performing scheduling scheme analysis based on the medical waste production prediction information and the working state information of the medical waste robot to determine a robot working scheduling scheme; and a whole-process management module for controlling the medical waste robot to perform medical waste treatment and whole-process recording according to the robot working scheduling scheme to obtain a robot medical waste treatment data chain, and performing whole-process traceability management through the robot medical waste treatment data chain.

[0008] The application provides one or more technical solutions, which have at least the following technical effects:

[0009] This application determines the data collection source by installing an IoT sensing device group on a medical waste robot. Next, the multi-dimensional data stream of medical waste is collected in real time, and technologies such as diversion identification, preprocessing, and anomaly recognition are used to obtain a standard multi-dimensional data stream of medical waste. Then, a time series neural network is trained using historical medical waste data sets to generate a medical waste prediction model and obtain medical waste production prediction information. After that, combined with the robot's working status information, the processing goals are defined and functions are constructed, and the ant colony algorithm is used to determine the robot's work scheduling plan. When controlling the robot to perform tasks, the road conditions are monitored in real time and the plan is corrected. The entire process is recorded to form a data chain for traceability and control, realizing the intelligent scheduling and management of medical waste treatment, achieving intelligent and automated medical waste treatment, improving treatment efficiency, reducing the occupational exposure risk of treatment personnel, and being able to reasonably schedule according to the actual situation of the waste, achieving the technical effect of refined management of the entire process.

[0010] The above content outlines the intelligent scheduling method and system of the present application for solving the medical waste robot based on the Internet of Things. The present application will describe the steps of the technical solution in detail in the following specific implementation methods to facilitate technical personnel to have a clear and complete understanding of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0012] Figure 1 This is a flow chart of the intelligent scheduling method of the medical waste robot based on the Internet of Things provided in an embodiment of the present application.

[0013] Figure 2 This is a structural diagram of the intelligent scheduling system of the medical waste robot based on the Internet of Things provided in an embodiment of the present application.

[0014] Explanation of the accompanying drawings: perception device installation module 1, data flow acquisition module 2, work information collection module 3, scheduling plan determination module 4, full-process management and control module 5. DETAILED DESCRIPTION

[0015] This application installs IoT sensing devices on medical waste robots to collect multidimensional data streams, which are then cleaned and processed to generate standardized data. A model trained using historical data predicts waste production information, and an algorithm is used to determine a scheduling plan based on the robot's operating status. The robot is controlled to execute tasks, road conditions are monitored in real time, and corrective measures are implemented. The entire process data chain is recorded for traceability and control, enabling intelligent scheduling and full-process management of medical waste disposal. This achieves intelligent and automated medical waste disposal, improves processing efficiency, reduces occupational exposure risks for waste handlers, and enables reasonable scheduling based on actual waste conditions, achieving the technical effect of refined management of the entire process.

[0016] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only some of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0017] It should be noted that any variations of the terms "include" and "have" are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or are inherent to these processes, methods, products or devices.

[0018] Example 1, as Figure 1 As shown, an intelligent scheduling method for medical waste robots based on the Internet of Things, wherein the method includes:

[0019] Step A100: Install an Internet of Things sensing device group on the medical waste robot, wherein the Internet of Things sensing device group includes a weight sensor, a type identification sensor, a GPS positioning device, an environmental sensor, and a video monitoring device.

[0020] Specifically, the equipment selection and hardware integration are carried out first. In view of the need to obtain weight, type, location, environmental status and visual images in real time in medical waste treatment scenarios, sensor equipment that meets medical environment standards is selected. The weight sensor uses a high-precision pressure sensor module with a range covering 0-50kg and an accuracy error controlled at ±0.5%, ensuring the accuracy of medical waste weighing data and providing a data basis for subsequent automatic billing functions; the type recognition sensor integrates a multi-spectral imaging module and an identification module, which can identify the 6 major categories and 23 minor categories of waste in the "Medical Waste Classification Catalog" and realize automatic classification through spectral reflectivity differences (such as the reflectivity characteristic value of infectious waste in the 450-550nm band is greater than 0.6); the GPS positioning device uses a centimeter-level differential positioning module, combined with an indoor UWB positioning system, to ensure the robot's positioning accuracy in the complex environment of the hospital. The robot's environmental sensor suite includes temperature and humidity monitoring modules, with a measurement range of 0-100% RH and -20°C-60°C, respectively, with an accuracy of ±2% RH and ±0.5°C. Its harmful gas detection sensitivity reaches ppb, providing real-time warnings of abnormal changes in the medical waste storage environment. The video surveillance system features a 1080P HD camera with infrared night vision and a wide-angle lens, supporting 360° scanning, a video frame rate of 25fps, and a storage resolution of 1920×1080, ensuring visual traceability of the waste collection site. To facilitate the robot's recognition of its surroundings, a lidar, depth camera, and ultrasonic sensor are also installed.

[0021] Next, the hardware installation layout was carried out, following the principle of "functional zoning and anti-interference design." A weight sensor was embedded in the load-bearing structure at the bottom of the robot's storage compartment, rigidly connected to the compartment body, and sensed pressure changes via strain gauges. A type identification sensor was installed above the compartment entrance, with its lens aimed vertically downward at the garbage disposal port to ensure positive imaging recognition during garbage placement. The GPS antenna was fixed to an unobstructed area on the top of the robot, and the environmental sensor group was integrated into the side of the fuselage in a ventilated position to avoid heat interference from mechanical components. The video surveillance camera was mounted on the front of the fuselage using an adjustable bracket, supporting remote control angle adjustment. Each sensor was connected to the robot's main control chip via interfaces such as RS-485 and USB-C, using shielded cables to reduce electromagnetic interference. Independent power modules were also configured to ensure stable power supply to the sensors.

[0022] Through the above steps, the IoT sensing device group and the medical waste robot achieve hardware physical integration, providing a reliable data foundation for subsequent data cleaning, waste production prediction and intelligent scheduling, while ensuring the realization of the self-weighing function.

[0023] Step A200: start the medical waste robot, and collect medical waste multi-dimensional data stream in real time through the Internet of Things sensing device group, perform data cleaning processing on the medical waste multi-dimensional data stream, and obtain standard medical waste multi-dimensional data stream.

[0024] In the embodiments of the present application, the medical waste multi-dimensional data stream refers to a multi-dimensional and multi-type data set collected in real time by the Internet of Things sensing device group installed on the medical waste robot.

[0025] Optionally, after the medical waste robot is started, first, the Internet of Things sensing device group collects multi-dimensional data in real time at a preset frequency (10 Hz), forming an original data stream containing weight, type, position, environmental parameters and video images. When each sensor works synchronously, the weight sensor feeds back the storage cabin load data in real time with an accuracy of ±0.5%, the type identification sensor generates 25 frames of garbage image data per second through the multi-spectral imaging module, the GPS positioning module outputs centimeter-level coordinates (positioning error ≤10 cm) in combination with the indoor UWB system, the environmental sensor group synchronously collects temperature and humidity (accuracy ±2% RH, ±0.5°C) and harmful gas concentration (sensitivity up to ppb level), and the video monitoring device records 1080P high-definition images at a frame rate of 25 fps. These original data are transmitted to the robot main control chip through the RS-485 / USB-C interface, and are initially packaged as binary data stream containing time stamps (single cycle data volume is about 1.2 MB).

[0026] Then, the medical waste multi-dimensional data stream collected by the Internet of Things sensing device group is identified according to the type, and multi-source medical waste data streams such as weight, type and positioning are obtained; the corresponding multi-source medical waste data preprocessing program is determined according to the characteristic information of each data stream; the multi-source data stream is processed through the preprocessing program to remove initial noise and format difference, and obtain available multi-source medical waste data stream; finally, the available data stream is identified and cleaned, and finally the standard medical waste multi-dimensional data stream is obtained, which is described in detail in A210-A240.

[0027] Step A300: predicting the production of garbage based on the standard medical waste multi-dimensional data stream, obtaining medical waste production prediction information, and collecting working state information of the medical waste robot.

[0028] In the embodiments of the present application, the garbage production prediction is to train a medical waste prediction model by collecting historical medical waste data set, analyze the cleaned standard medical waste multi-dimensional data stream based on the model, and output prediction information such as medical waste production, type ratio and peak period of each department in the future period.

[0029] In one embodiment of the present application, first, a historical medical waste data set is collected and time-series divided and marked to obtain a historical medical waste sequence data set, and then a time series neural network is used to train and generate a medical waste prediction model; based on the prediction model, the standard medical waste multidimensional data stream is analyzed to obtain medical waste production prediction information. The specific steps are described in detail in A310-A330.

[0030] While predicting waste production, the robot's built-in sensors collect real-time operating status information. At the hardware level, a power sensor monitors the remaining charge in the lithium battery (with an accuracy of ±1.5%), triggering a low-battery warning when the charge falls below 20%. An odometer records the robot's distance traveled (with an error of ≤0.5%) and calculates its current speed (with a resolution of 0.1 m / s) using GPS positioning data. A gyroscope and accelerometer sense the robot's posture (with an inclination error of ≤2°) to determine whether a collision or tipping has occurred. A task status register provides real-time feedback on the current task type (e.g., "Go to the 3rd floor laboratory to collect chemical waste"), completion progress (percentage), and fault codes (e.g., 0x01 indicates a stuck sorting hatch).

[0031] At the software level, status data is packaged (including timestamp, robot ID, and status parameter array) by the edge computing module (a distributed computing unit deployed locally on the medical waste robot or near its communication network, capable of real-time data processing and edge intelligent decision-making, and assuming the core function of data preprocessing in the intelligent scheduling system, specifically described in detail in A210-A240). It is uploaded to the cloud via the MQTT protocol at a frequency of 5Hz (transmission delay ≤150ms), forming a work status information flow (single data size is approximately 500B) covering the four dimensions of "power-location-task-fault". The system verifies the status data in real time. When it detects that there is no update on the task progress or the speed is abnormal (>1.5m / s) for three consecutive times, the robot self-test program is automatically triggered to ensure the real-time and accuracy of the work status information.

[0032] Through the above steps, a closed loop of "historical data training-real-time production prediction-multi-dimensional status monitoring" is realized, providing dual data support for intelligent scheduling, that is, medical waste prediction information is used for forward-looking planning of collection routes and frequencies, and robot working status information ensures the feasibility of dynamic adjustment of scheduling plans.

[0033] Step A400: Perform a scheduling plan analysis based on the medical waste production prediction information and the working status information of the medical waste robot to determine the robot work scheduling plan.

[0034] Specifically, define the robot's waste disposal goals and construct an objective function. Using the medical waste production prediction information and the robot's working status information as constraints, obtain the scheduling plan space through ant colony algorithm analysis, and then use the objective function to globally optimize within this space to determine the robot's work scheduling plan. The specific steps are detailed in A410-A430.

[0035] Step A500: Control the medical waste robot to perform medical waste treatment and full-process recording according to the robot work scheduling plan, obtain the robot medical waste treatment data chain, and perform full-process traceability control through the robot medical waste treatment data chain.

[0036] Specifically, first, control the medical waste robot to perform medical waste treatment according to the scheduling plan, monitor the working road conditions in real time and identify obstacle information, and correct the scheduling plan based on the information to obtain the robot work scheduling correction plan. The specific steps are described in detail in A510-A530.

[0037] During the entire medical waste disposal process, the robot will record the entire process. The robot is equipped with a full range of monitoring cameras and data recording modules, which record real-time video of each link such as waste collection, transportation, and storage with a high-definition image quality of 1080P or above and a frame rate of more than 25 frames per second. At the same time, the data recording module will automatically record the timestamp, location information (accurate to the specific department and coordinates) and related task parameters (such as the type and weight of collected waste, etc.) of each operation step. These data will be stored in an encrypted form in the robot's local storage device. The storage capacity must meet the continuous recording requirements of at least 30 days. At the same time, it will be uploaded to the cloud server in real time through wireless communication technology for backup and further processing.

[0038] Based on the data recorded throughout the entire process, a robotic medical waste disposal data chain is generated. This chain integrates all key information from the source of waste generation (department information) to the final treatment destination (waste treatment center), forming a complete and traceable information chain. Each data node in the chain contains detailed metadata, such as time, location, operation type, and responsible person, to ensure data integrity and accuracy.

[0039] Finally, the robotic medical waste treatment data chain is used to conduct full-process traceability control. Technical personnel in this field can use a dedicated management platform to input specific query conditions (such as time range, department name, waste type, etc.) to quickly locate and view relevant medical waste treatment records. For example, when an abnormality is found in a batch of medical waste, the data chain can be used to trace back to detailed information such as the department where the waste was generated, the collection time, the transportation route, and the treatment method, so as to promptly troubleshoot the problem and take appropriate measures. At the same time, the data chain can also provide strong data support for quality assessment, cost accounting, compliance inspection, etc. of medical waste treatment, thereby achieving refined management and effective control of the entire process of medical waste treatment.

[0040] Furthermore, step A200 in the method provided in the embodiment of the present application includes:

[0041] A210: Divert and identify the multi-dimensional data stream of medical waste according to the IoT sensing device group to obtain a multi-source medical waste data stream.

[0042] A220: Determine a multi-source medical waste data preprocessing procedure based on the characteristic information of the multi-source medical waste data stream.

[0043] A230: Preprocessing the multi-source medical waste data stream based on the multi-source medical waste data preprocessing program to obtain a usable multi-source medical waste data stream.

[0044] A240: Perform anomaly identification and data cleaning on the available multi-source medical waste data stream to obtain a standard medical waste multi-dimensional data stream.

[0045] Specifically, after the medical waste robot is started, the IoT sensing device group collects multi-dimensional data streams of medical waste in real time at a frequency of 10Hz. After these data are transmitted to the robot's main control chip via the RS-485 / USB-C interface, they are first diverted and identified: a label is added to each data according to the sensor's unique ID (such as weight sensor ID-W01, type identification sensor ID-T02), and the data is split into five independent sub-data streams: weight, type, positioning, environment, and image.

[0046] At the same time, a CRC-16 checksum algorithm is used to eliminate transmission errors: First, the sensor adds a 16-bit checksum to the original data frame (such as sensor output signals such as weight and type). This checksum is generated by performing a modulo-2 division operation on the data frame using a preset generating polynomial (such as CRC-16-CCITT). When the main control chip receives the data frame, it extracts the data portion and recalculates the CRC checksum value, comparing it bit by bit with the checksum carried at the end of the frame. If the two do not match, the data frame is deemed to have been transmitted incorrectly (i.e., the checksum failed). The number of frames with consecutive checksum failures is counted. If three consecutive frames fail the checksum failure, these abnormal data frames are marked as red and pending, triggering a retransmission mechanism or manual intervention to ensure that the checksum pass rate of data frames entering the subsequent processing flow is ≥99.9%. This effectively eliminates erroneous data caused by electromagnetic interference and transmission errors, thereby forming a preliminarily classified multi-source medical waste data stream.

[0047] Next, based on the characteristic differences of the above five types of sub-data streams, the edge computing module built into the main control chip automatically matches the preprocessing program:

[0048] Step a: Weight data is fed into a Kalman filter (50ms iteration cycle) to filter out high-frequency noise generated by mechanical vibration. Specifically, the algorithm predicts the current weight estimate based on the previous weight and the robot's motion state (speed and acceleration). The sensor's measured value and the predicted value are then weighted and fused. The weights are adaptively adjusted using the noise covariance matrix to filter out vibration noise with a frequency greater than 10Hz (such as interference from cart jolting). This processing reduces the noise amplitude of the weight data, providing stable input data for waste weighing and load balancing calculations.

[0049] Step b: The type image data is processed using a combination of histogram equalization and median filtering to enhance the contrast of garbage surface features and is then cut to a standard input size of 224×224 pixels. Specifically, histogram equalization expands the dynamic range of the image's grayscale values ​​(e.g., from [50,150] to [0,255]), enhancing the contrast of garbage surface textures (e.g., reflective markings on IV bags and the metallic sheen of needles). The median filter algorithm addresses salt and pepper noise (e.g., sensor noise) in the image by replacing the center pixel with the neighborhood median using a 3×3 pixel sliding window to remove discrete noise points. The processed image is further cut to a standard size of 224×224 pixels to match the input requirements of the subsequent classification model. This ensures that each image contains the complete garbage feature area (e.g., a garbage sample directly in the center of the drop-off port), improving the confidence of the type recognition algorithm.

[0050] Step c: The location data undergoes coordinate conversion based on the hospital's electronic map (WGS84 geographic coordinate system to local coordinate system, with an error of ≤5cm) and is associated with floor and department information (e.g., "3rd floor Laboratory Department" corresponds to X, Y coordinates). This conversion process first converts longitude and latitude into rectangular coordinates using a seven-parameter model (translation, rotation, and scaling parameters). A secondary calibration is then performed using the hospital's electronic map's coordinate offset (e.g., the 1st floor reference point coordinates are X = 0cm, Y = 0cm). The final error is kept within 5cm (meeting the navigation accuracy requirements in narrow corridors). Simultaneously, the system pre-generates a floor-to-department coordinate mapping table (e.g., the 3rd floor Laboratory Department corresponds to coordinates within the range X = 12000-13000cm, Y = 8000-9000cm). Using spatial geometry algorithms (e.g., point-to-polygon containment detection), the system automatically associates the current coordinates with the department name. This transforms the location data from simple numerical coordinates into meaningful information such as "3rd floor Laboratory Department," supporting subsequent department-by-department waste generation statistics and task allocation.

[0051] Step d: The environmental data and image streams are synchronized using a hardware clock. (Due to different sampling frequencies (10 Hz for the former and 25 fps for the latter), hardware clock synchronization is required to ensure temporal consistency between the two data types.) Specifically, the edge computing unit incorporates a high-precision crystal oscillator (with a clock error of ±5 ppm) or accesses a GPS timing signal to assign a unified timestamp reference to each sensor and camera. When the environmental sensor collects data (e.g., 14:00:00.000), the video stream simultaneously records the frames within 40 ms before and after that moment (corresponding to one frame at 25 fps). Timestamp matching (allowing for ±50 ms error) links the environmental parameters of "25°C, 60% RH" to the waste disposal imagery of the same moment (e.g., footage showing medical staff disposing of infectious waste at 14:00:00.030), creating temporally and spatially aligned linked data. This synchronization ensures that subsequent traceability can quickly locate the corresponding imagery when an environmental anomaly (e.g., a sudden temperature rise) is detected, providing accurate time-series data support for analyzing the correlation between the medical waste storage environment and safety risks.

[0052] The above preprocessed data are uniformly encapsulated in JSON format (a single data entry is about 2KB), including fields such as "timestamp, sensor ID, data value, device status", forming a usable multi-source medical waste data stream with standardized structure.

[0053] Finally, the available multi-source medical waste data stream is analyzed for abnormal thresholds to determine the abnormal judgment threshold. Based on this threshold, the abnormal multi-source medical waste data set is judged and cleaned, and finally a standard medical waste multi-dimensional data stream is obtained. The specific steps are described in detail in A241-A243.

[0054] Through the above steps, high-quality input data with controllable errors and unified format is provided for subsequent garbage production predictions, which solves the fusion of multi-source heterogeneous data and supports the accurate implementation of automatic recording and full-process traceability functions.

[0055] Furthermore, step A240 in the method provided in the embodiment of the present application includes:

[0056] A241: Perform abnormal threshold analysis on the available multi-source medical waste data streams respectively to determine the abnormality judgment threshold of the multi-source medical waste data.

[0057] A242: Based on the multi-source medical waste data abnormality judgment threshold, abnormal data judgment is performed on the available multi-source medical waste data stream to obtain an abnormal multi-source medical waste data set.

[0058] A243: Perform data cleaning on the abnormal multi-source medical waste data set to obtain a standard medical waste multi-dimensional data stream.

[0059] In the embodiment of the present application, abnormal threshold analysis is a step of setting reasonable abnormality judgment standards for each dimension of data by analyzing data characteristics and combining historical statistical laws and industry standards for the pre-processed multi-source medical waste data stream.

[0060] Optionally, first, analyze the abnormal thresholds for data in different dimensions, such as weight, type, location, and environment. For weight data, set a physical range threshold (0-50kg, matching the maximum load capacity of the storage compartment). At the same time, calculate the dynamic fluctuation threshold based on historical data. That is, take the 95% percentile of the weight change rate at adjacent moments in the past 7 days (usually 12%-15%). When the real-time change rate is greater than 15%, trigger an abnormal warning.

[0061] The type identification data is based on the confidence distribution during classification model training. The confidence threshold is set at 0.9, and data below this value is marked as fuzzy data. The specific training process of the classification model is as follows:

[0062] Step e: Obtain multispectral image data of a large amount of medical waste after preprocessing and segmentation (detailed description in step b), label these preprocessed image data with corresponding waste type labels to form a training data set.

[0063] Step f: Select an appropriate deep learning architecture, such as ResNet-18, as the base model. Input the training dataset into the model and continuously adjust the model parameters using the backpropagation algorithm to minimize the error between the predicted results and the labeled labels. During training, use methods such as cross-validation to evaluate and optimize the model to prevent overfitting. Training continues for multiple rounds until the model performance reaches a stable level.

[0064] Step g: Finally, a stable medical waste classification model is obtained. The input of this classification model is a preprocessed 224×224 pixel multispectral image of medical waste. The output is the medical waste type corresponding to the image and the confidence level of the classification result. For example, the output may be "Infectious waste, confidence level 95%."

[0065] The positioning data is combined with the hospital's floor plan to define 20 restricted areas (such as operating rooms and sterile pharmacies, corresponding to the electronic map coordinate set), and a moving speed threshold of 1.5m / s is set (exceeding this value may cause collision risks). Environmental data is referenced in the "Medical Waste Management Regulations" to set safety intervals, such as temperature 2-40°C, humidity 30%-80% RH, and formaldehyde concentration ≤0.08mg / m 3 , if it exceeds the range, it is judged as an environmental anomaly. The integration of the above thresholds is the abnormality judgment threshold for multi-source medical waste data.

[0066] Next, based on the above threshold, the edge computing module scans the available multi-source data streams point by point at a frequency of 10Hz, and performs abnormal data marking through the logic judgment engine. Assuming that the weight data shows 55kg (out of range) or a sudden increase from 10kg to 25kg within 10 seconds (change rate 150%>15%), a weight abnormality is triggered; the type recognition result is "unclassified" and the confidence level is 0.85 (<0.9), and it is marked as type fuzzy; the positioning coordinates fall into the forbidden area (such as the operating room coordinate range X=5000-6000cm, Y=3000-4000cm) or the speed is 1.8m / s (>1.5m / s), which is determined to be a positioning abnormality; the temperature in the environmental data is 45℃ (>40℃) or formaldehyde is 0.1mg / m 3 (>0.08mg / m 3 ) and recorded as an environmental violation. Data points that meet any of the conditions are extracted into the abnormal multi-source medical waste dataset, which typically accounts for 1.5%-3% of the preprocessed data and includes four main types of anomalies: value violation, classification ambiguity, location conflict, and environmental risk.

[0067] Finally, for abnormal data sets, the system initiates a hierarchical cleaning mechanism. For numerical anomalies such as weight, temperature and humidity, the sliding window interpolation method is preferentially used. If there is a valid value within 5 cycles (500ms) before and after the abnormal point, it is repaired through linear interpolation (for example, when weight data is missing, the average value of 12kg at the previous moment and 13kg at the next moment is taken as 12.5kg), and the repair success rate reaches 96.2%; if there is no valid value, it is marked as "N / A" and the sensor self-test is triggered (fault diagnosis is completed within 2 minutes).

[0068] Type-ambiguous data enters the secondary recognition process, and the edge computing module uploads the image data to the cloud-based enhanced model (processing delay ≤ 300ms under computing power support). After review, samples with a confidence level still less than 0.9 are automatically associated with the video clips with the corresponding timestamps (1 second before and after, 25 frames of image), and a manual review work order is generated.

[0069] Anomaly data is corrected using a trajectory fitting algorithm. A Bezier curve is constructed based on valid coordinates (≥3) within the last 30 seconds to infer the reasonable coordinates of the anomaly point (with an error ≤ 20 cm), ensuring path continuity (for example, coordinate jumps caused by elevator signal loss can improve trajectory smoothness by 40% after fitting). Format standardization is performed simultaneously during the cleaning process. Weight data is uniformly retained to two decimal places, type data is mapped to the Medical Waste Classification Catalog code (such as HW01 for infectious waste), positioning coordinates are converted to the hospital's local coordinate system (in cm), and environmental parameters are output according to national standard units (such as °C and %RH).

[0070] The final generated standard medical waste multi-dimensional data flow is uploaded to the cloud through the MQTT protocol. A single piece of data contains multiple standardized fields (time, department, type code, weight, temperature and humidity, coordinates, etc.).

[0071] Through the above steps, a high-quality data foundation is laid for the subsequent garbage production prediction model training and intelligent scheduling algorithm operation, and the key conversion from raw heterogeneous data to standardized usable data is achieved, supporting the realization of the core functions of full-process traceability control, intelligent scheduling optimization and automatic classification.

[0072] Furthermore, step A300 in the method provided in the embodiment of the present application includes:

[0073] A310: Collect and obtain a historical medical waste dataset, perform time series division and identification on the historical medical waste dataset, and obtain a historical medical waste sequence dataset.

[0074] A320: Use a time series neural network to perform model training on the historical medical waste sequence dataset to generate a medical waste prediction model.

[0075] A330: Perform waste production prediction on the standard medical waste multidimensional data stream based on the medical waste prediction model to obtain the medical waste production prediction information.

[0076] In an embodiment of the present application, a time series neural network is a neural network model used to process data with time dependencies.

[0077] Specifically, when obtaining forecast information on medical waste production, we first build a historical data foundation. We extract a 12-month historical medical waste dataset from a cloud database (including fields such as collection time, generating department, waste weight, type distribution, and collection period). Using a time series partitioning algorithm, we segment the data into a "year-quarter-month-day-hour" granularity, forming a historical medical waste sequence dataset with one hour as the smallest time unit.

[0078] Based on the above dataset, an LSTM time series neural network is used for model training to generate a medical waste prediction model with multi-task prediction capabilities. The specific steps are described in detail in A321-A323.

[0079] After the model is deployed, a rolling forecast is performed on the standard multidimensional data stream of medical waste input in real time. Using the current time as the base point, a sliding window of data from the previous 24 hours is captured. Numerical features (such as weight, temperature, and humidity) are scaled to the [-1, 1] range using the Z-score normalization method. Categorical features (such as department and type code) are converted to a one-hot encoding (for a categorical variable with n different categories, one-hot encoding creates a binary vector of length n, with only one element set to 1 and all other elements set to 0, indicating the category to which the variable belongs). This is then fed into the forecasting model.

[0080] The model outputs forecast information for the next four hours, including: hourly waste generation forecasts for each department (e.g., the emergency department is expected to generate 18.5kg±1.5kg of hazardous waste from 15:00-16:00); dynamic changes in type distribution ratios (e.g., the proportion of chemical waste during the afternoon surgery peak increases from 18% to 25%); and peak period warnings (three peak periods are identified in the next 24 hours, with probabilities of 93%, 89%, and 91%, respectively).

[0081] The prediction results are cross-validated with the hospital HIS system's patient volume data and operating room schedule. When the prediction confidence is less than 85%, a secondary prediction is triggered (calling the data of the same time period in the past 7 days to enhance the model input) to ensure that the error rate of the output medical waste production prediction information is reduced.

[0082] Through the above steps, high-precision data support is provided for the forward-looking planning of subsequent robot scheduling solutions.

[0083] Furthermore, step A320 in the method provided in the embodiment of the present application includes:

[0084] A321: Use a time series neural network to perform attribute identification training on the historical medical waste sequence dataset to obtain a waste generation volume prediction model, a waste type distribution prediction model, and a waste generation peak period prediction model.

[0085] A322: The garbage generation amount prediction model, the garbage type distribution prediction model, and the garbage generation peak period prediction model are combined to obtain a basic garbage prediction model.

[0086] A323: Perform performance verification and optimization on the basic waste prediction model to generate the medical waste prediction model.

[0087] In the embodiment of the present application, attribute identification training is a training method for a data set with multi-attribute characteristics, which allows the model to learn the relationship between different attributes and target predictions.

[0088] Specifically, first, a multi-task sub-model system was constructed using attribute identification training. For a historical medical waste sequence dataset (with a time resolution of 1 hour), a multi-output time series neural network architecture based on LSTM was used to divide the input features into three attribute spaces: time attributes (hourly codes 0-23, weekday / weekend identifiers, and seasonal factors), department attributes (60 departments were one-hot encoded and then reduced to 5 dimensions using PCA), and time series attributes (mean output of the previous 3 hours, type distribution variance, and collection frequency fluctuation coefficient). The network was designed with three independent output branches: the output prediction branch outputs continuous values ​​(unit: kg / h) using the mean squared error (MSE) loss function; the type distribution branch outputs the probability distribution of 6 types of waste (summed to 1) using a categorical cross-entropy loss; and the peak period branch outputs a binary classification result (peak / off-peak, threshold 0.8) using a binary cross-entropy loss.

[0089] During training, the first two LSTM hidden layers (each with 128 memory cells and a dropout rate of 0.2) are shared to extract common temporal features. The third hidden layer then performs feature mapping specific to each task branch. After 100 rounds of iterative training (batch size 32, Adam optimizer learning rate 0.001), the generation volume prediction error stabilized at ±8%, the type distribution prediction accuracy exceeded 90%, and the peak hour identification F1 score exceeded 0.9, thus forming independent prediction sub-models for garbage generation volume, type distribution, and peak hour.

[0090] Next, the model merging process begins, integrating the network structures of the three sub-models in a lightweight manner: retaining the shared first two feature extraction layers, and concatenating the output layers of each task branch into a unified output interface, allowing a single input to simultaneously generate three types of prediction results (production quantity, type probability vector, and peak probability value). The merged basic garbage prediction model maintains a 15-dimensional input dimension (5-dimensional time + 5-dimensional department + 5-dimensional time series), while the output dimension is expanded to 8 dimensions (1-dimensional production + 6-dimensional type probability + 1-dimensional peak probability). A confidence calibration layer is added to the model output to calibrate the peak period prediction probability to a credible interval of [0.7, 0.95] to ensure the numerical stability of the multi-task output. Testing on a cross-validation set (accounting for 20% of the dataset) shows that the merged model (basic garbage prediction model) achieves improved inference speed compared to the independent sub-models, and the performance degradation of each sub-task is controlled within 3%, meeting the computing power constraints of edge computing nodes.

[0091] Finally, performance verification and optimization were conducted using a dual process of "historical data backtesting and real-time scenario stress testing." First, using the past three months of historical data, the model's predicted values ​​were compared with the actual values. The calculated mean absolute error (MAE) of the generated volume forecast was 1.2 kg / h, the Kullback-Leibler divergence (KL divergence) of the type distribution remained stable below 0.15, the missed detection rate during peak hours (actual peak hours not identified) was ≤5%, and the false alarm rate (non-peak hours misidentified) was ≤3%. For the weaker early morning hours (where garbage generation is low and fluctuating), data augmentation techniques were used to expand the training sample (copying the data from this period and adding ±10% random noise). The LSTM forget gate parameter was adjusted (increased from 0.8 to 0.85) to strengthen long-term memory capabilities. In a real-world hospital deployment, three departments were selected for 72-hour real-time validation. When patient volume surged (e.g., the emergency department saw over 500 patients per day, a significant change from the typical daily volume), the model's prediction error temporarily increased from the typical 8% to 12%. This triggered a dynamic optimization mechanism, automatically adding data from similar abnormal scenarios from the previous seven days to the training set. After five rounds of rapid incremental training, the error dropped back to less than 9%. The resulting medical waste prediction model possesses multi-dimensional prediction capabilities (simultaneously outputting information on volume, type, and peak volume). Over 90% of the data had a prediction confidence level of 85% or higher, providing highly accurate forward-looking data support for intelligent robotic scheduling.

[0092] Furthermore, step A400 in the method provided in the embodiment of the present application includes:

[0093] A410: Define a robot garbage disposal goal, and construct a robot garbage disposal objective function based on the robot garbage disposal goal.

[0094] A420: Using the medical waste production prediction information and the working status information of the medical waste robot as constraints, an ant colony algorithm is used to perform a scheduling solution analysis on the constraints to obtain a robot scheduling solution space.

[0095] A430: Utilize the robot garbage disposal objective function to perform global optimization in the robot scheduling plan space to determine the robot work scheduling plan.

[0096] In the embodiment of the present application, the waste disposal objective function is used to quantify the optimization goal in the intelligent scheduling of medical waste robots based on the Internet of Things. The robot scheduling solution space refers to the set of all possible robot scheduling solutions under given constraints.

[0097] Specifically, first, clarify the treatment objectives and construct a mathematical model. Based on the actual needs of hospital medical waste treatment, define the robot waste treatment objectives, such as minimizing the total time spent on waste collection (to ensure timeliness and avoid infection risks), maximizing resource utilization (reducing robot idle mileage and reducing energy consumption), and balancing the load of each robot (preventing a robot from overworking and causing an increase in failure rate). Based on these objectives, construct the robot waste treatment objective function. Taking minimizing the total time spent on waste collection as an example, its objective function can be expressed as: Where n is the number of medical waste collection points (e.g. if there are 60 departments in a hospital, n is 60), m is the number of robots, and t ij represents the travel time of robot j from the current position to the collection point i, x ij is a 0-1 variable (1 means robot j goes to collection point i, 0 means it does not go). At the same time, those skilled in the art integrate other goals (resource utilization, load balancing) into the objective function through weighted coefficients to form a comprehensive optimization goal.

[0098] Then, the medical waste production prediction information and the robot working state information are taken as constraint conditions, and an ant colony algorithm is used for analysis. The medical waste production prediction information includes the amount of waste produced by each department in the next 4 hours, the type distribution, and the peak period (for example, the emergency department is expected to produce 18.5 kg of injury waste from 15:00 to 16:00), and the robot working state information includes the current position coordinates, the remaining capacity of the storage cabin, the battery power, and the like. The ant colony algorithm simulates the foraging process of ants, takes the scheduling path of each robot as the walking path of an ant, and searches for feasible solutions in the robot scheduling scheme space. The algorithm iteratively calculates different scheduling schemes (such as robot 1 responsible for departments 1-20, robot 2 responsible for departments 21-40, and the like) through the pheromone updating mechanism (the initial pheromone concentration is set to 1) and the state transition probability formula (combined with the objective function and the constraint conditions), and generates a robot scheduling scheme space containing more than 1000 feasible schemes after 50 iterations (the convergence speed is usually ≤10 minutes).

[0099] Finally, the robot waste treatment objective function is used for global optimization in the scheme space. Each scheme in the robot scheduling scheme space is substituted into the objective function for calculation and evaluation. For example, the total waste collection time of scheme 1 is 60 minutes, the resource utilization rate is 85%, the load balancing coefficient is 0.9, and the comprehensive score is 88; the corresponding indicators of scheme 2 are 55 minutes, 90%, 0.85, and the comprehensive score is 92. By comparing the scores of all schemes, the scheme with the highest score is selected as the final robot working scheduling scheme.

[0100] Through the above steps, the robot working scheduling scheme with the highest score is obtained, and efficient, balanced and energy-saving medical waste treatment is effectively realized.

[0101] Further, the step A500 in the method provided in the embodiment of the application comprises:

[0102] A510: controlling the medical waste robot to perform medical waste treatment according to the robot working scheduling scheme, and monitoring and acquiring the robot working road condition in real time.

[0103] A520: performing obstacle identification on the robot working road condition to determine road obstacle information.

[0104] A530: modifying the robot working scheduling scheme based on the road obstacle information to obtain a robot working scheduling modification scheme.

[0105] In one embodiment, when the medical waste robot is executing medical waste disposal according to the work scheduling scheme, first, let the robot start action according to the established work scheduling scheme, while monitoring its work road condition in real time. The medical waste robot is equipped with various sensors, such as laser radar, camera and ultrasonic sensor, etc., for collecting road condition information. Laser radar scans the surrounding environment at a frequency of 100,000-200,000 data points per second, which can accurately measure the distance, and the measurement accuracy can reach centimeter level; the camera captures images at a speed of 30-60 frames per second, providing data for subsequent visual analysis; the ultrasonic sensor can assist in detecting obstacles at close range. The data of these sensors will be transmitted to the control system of the robot in real time, and the control system will integrate and process these data to form a preliminary understanding of the current work road condition.

[0106] Next, the obtained robot work road condition is identified for obstacles to determine the road condition obstacle information. For laser radar data, a point cloud processing-based obstacle identification algorithm is used. The algorithm first filters the original point cloud data to remove noise points and outliers to improve data quality. Then, the point cloud data is divided into different clusters by clustering algorithm, and each cluster represents a possible obstacle. Common clustering algorithms such as DBSCAN (Density-Based Spatial Clustering Algorithm) can adapt to obstacles of different shapes and sizes by clustering based on the density of point clouds. For images captured by the camera, deep learning object detection algorithms such as YOLO (You Only Look Once) series algorithms are used. YOLO algorithm has high real-time performance, which can detect obstacles in images within a short time and give their position and category information. The identification results of laser radar and camera are fused to improve the accuracy and reliability of obstacle identification.

[0107] Then, the robot work scheduling scheme is corrected based on the determined road condition obstacle information. A* algorithm is used for path planning correction. A* algorithm considers both the actual cost and the estimated cost from the current position to the target position. First, the current position and target position of the robot are taken as input, and the road condition obstacle information is converted into obstacle regions on the map. Then, A* algorithm searches for an optimal path on the map that avoids obstacles. During the search process, the algorithm continuously calculates the cost of each node, selects the node with the smallest cost for expansion, and continues until the target position is found or all possible nodes are traversed. According to the newly planned path, the original robot work scheduling scheme is adjusted to form a robot work scheduling correction scheme. For example, if the original scheme requires the robot to pass through an area blocked by obstacles, the correction scheme will let the robot choose a path around the area to ensure that the robot can safely and efficiently complete the medical waste disposal task.

[0108] Finally, after the robot completes the medical waste disposal task, the edge computing module initiates the automatic billing process. First, the robot uses its built-in sensors to collect and upload key data, such as the weight and volume of collected medical waste, mileage traveled, and operating hours. After receiving this data, the module performs calculations based on billing rules pre-defined by technical personnel. For waste collection, different unit prices are set for each type of waste, and fees are calculated based on the weight or volume collected. Corresponding charging standards are also set for the robot's mileage and operating hours. The module then aggregates all costs to determine the total cost of the medical waste disposal. Finally, a detailed expense list is generated, including the detailed and total amounts for each fee, and sent to the relevant management department or fee settlement party, completing the automatic billing process.

[0109] By real-time monitoring of road conditions and identifying obstacle information, the robot's work scheduling plan is dynamically corrected based on this, achieving intelligent obstacle avoidance of the robot's path and adaptive adjustment of the scheduling plan during medical waste treatment, ensuring the realization of the automatic billing function, and effectively improving the safety of robot operations and the technical effect of task execution efficiency.

[0110] In summary, the intelligent scheduling method for medical waste robots based on the Internet of Things provided in the embodiments of the present application has the following technical effects:

[0111] This application constructs a data transmission channel between the medical waste data preprocessing module and the intelligent scheduling decision module, uses an anomaly threshold analysis algorithm to identify data anomalies, and obtains standardized medical waste multi-dimensional data streams through operations such as Kalman filtering denoising and histogram equalization enhancement. In the prediction model training module, time series division and attribute identification training are carried out. After LSTM network iterative optimization and multi-model merging verification, combined with ant colony algorithm optimization and path dynamic correction mechanisms, the robot work scheduling plan is generated and adjusted based on real-time road conditions and task requirements to ensure efficient and orderly medical waste treatment, achieve intelligent and automated medical waste treatment, improve treatment efficiency, reduce the occupational exposure risk of treatment personnel, and can reasonably schedule according to the actual situation of the waste, achieving the technical effect of refined management of the entire process.

[0112] Example 2, as Figure 2 As shown, based on the same inventive concept as the aforementioned embodiment 1, the embodiment of the present application provides an intelligent scheduling system for medical waste robots based on the Internet of Things, the system comprising:

[0113] The sensing device installation module 1 is used to install an Internet of Things sensing device group on the medical waste robot. The Internet of Things sensing device group includes a weight sensor, a type identification sensor, a GPS positioning device, an environmental sensor and a video monitoring device.

[0114] The data stream acquisition module 2 is used to start the medical waste robot and collect and obtain the medical waste multidimensional data stream in real time through the Internet of Things sensing device group, perform data cleaning processing on the medical waste multidimensional data stream, and obtain a standard medical waste multidimensional data stream.

[0115] The working information collection module 3 is used to perform waste production prediction on the standard medical waste multi-dimensional data stream, obtain medical waste production prediction information, and simultaneously collect and obtain the working status information of the medical waste robot.

[0116] The scheduling scheme determination module 4 is used to perform scheduling scheme analysis based on the medical waste production prediction information and the working status information of the medical waste robot to determine the robot work scheduling scheme.

[0117] The full-process control module 5 is used to control the medical waste robot to perform medical waste treatment and full-process recording according to the robot work scheduling plan, obtain the robot medical waste treatment data chain, and perform full-process traceability control through the robot medical waste treatment data chain.

[0118] Furthermore, the data stream obtaining module 2 is configured to perform the following steps:

[0119] The multi-dimensional data stream of medical waste is diverted and identified according to the IoT sensing device group to obtain a multi-source medical waste data stream; a multi-source medical waste data preprocessing program is determined based on the characteristic information of the multi-source medical waste data stream; the multi-source medical waste data stream is preprocessed based on the multi-source medical waste data preprocessing program to obtain a usable multi-source medical waste data stream; anomaly identification and data cleaning are performed on the usable multi-source medical waste data stream to obtain a standard medical waste multi-dimensional data stream.

[0120] Furthermore, the data stream obtaining module 2 is configured to perform the following steps:

[0121] The available multi-source medical waste data streams are respectively analyzed by abnormal threshold values ​​to determine the abnormal judgment threshold values ​​of the multi-source medical waste data; abnormal data judgment is performed on the available multi-source medical waste data streams based on the multi-source medical waste data abnormal judgment threshold values ​​to obtain an abnormal multi-source medical waste data set; data cleaning is performed on the abnormal multi-source medical waste data set to obtain a standard medical waste multi-dimensional data stream.

[0122] Furthermore, the work information collection module 3 is used to perform the following steps:

[0123] A historical medical waste data set is collected and acquired, and the historical medical waste data set is divided and labeled into time series to obtain a historical medical waste sequence data set; a time series neural network is used to perform model training on the historical medical waste sequence data set to generate a medical waste prediction model; and based on the medical waste prediction model, waste production prediction is performed on the standard medical waste multidimensional data stream to obtain the medical waste production prediction information.

[0124] Furthermore, the work information collection module 3 is used to perform the following steps:

[0125] A time series neural network is used to perform attribute identification training on the historical medical waste sequence data set to obtain a waste generation amount prediction model, a waste type distribution prediction model, and a waste generation peak period prediction model; the waste generation amount prediction model, the waste type distribution prediction model, and the waste generation peak period prediction model are merged to obtain a basic waste prediction model; the basic waste prediction model is performance verified and optimized to generate the medical waste prediction model.

[0126] Furthermore, the scheduling scheme determination module 4 is configured to perform the following steps:

[0127] The robot waste disposal goal is defined, and a robot waste disposal objective function is constructed based on the robot waste disposal goal; the medical waste production prediction information and the working status information of the medical waste robot are used as constraints, and an ant colony algorithm is used to perform a scheduling plan analysis on the constraints to obtain a robot scheduling plan space; the robot waste disposal objective function is used to perform a global optimization in the robot scheduling plan space to determine the robot work scheduling plan.

[0128] Furthermore, the full-process control module 5 is used to perform the following steps:

[0129] The medical waste robot is controlled to perform medical waste treatment according to the robot work scheduling plan, and the robot working road condition is monitored in real time; obstacles are identified on the robot working road condition to determine road condition obstacle information; the robot work scheduling plan is corrected based on the road condition obstacle information to obtain a robot work scheduling correction plan.

[0130] The intelligent scheduling system of the medical waste robot based on the Internet of Things provided by the embodiment of the present invention can execute the intelligent scheduling method of the medical waste robot based on the Internet of Things provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0131] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other and are not used to limit the scope of protection of the present invention.

[0132] The above specific embodiments do not constitute a limitation to the scope of protection of this application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements and improvements made within the spirit and principles of this application should be included in the scope of protection of this application. In some cases, the actions or steps recorded in this application can be performed in an order different from that in the embodiments and can still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. An intelligent dispatching method for medical waste robots based on the Internet of Things, characterized in that: The method comprises: Install an Internet of Things sensing device group on the medical waste robot, which includes a weight sensor, a type identification sensor, a GPS positioning device, an environmental sensor, and a video monitoring device; Starting the medical waste robot, and acquiring a medical waste multidimensional data stream in real time through the IoT sensing device group, performing data cleaning on the medical waste multidimensional data stream to obtain a standard medical waste multidimensional data stream; Performing waste production prediction on the standard medical waste multi-dimensional data stream to obtain medical waste production prediction information, and simultaneously collecting and obtaining working status information of the medical waste robot; Performing a scheduling analysis based on the medical waste production forecast information and the working status information of the medical waste robot to determine a robot work scheduling plan; The medical waste robot is controlled to perform medical waste treatment and full-process recording according to the robot work scheduling plan, obtain a robot medical waste treatment data chain, and perform full-process traceability control through the robot medical waste treatment data chain.

2. The intelligent scheduling method of medical waste robot based on the Internet of Things according to claim 1, characterized in that: The method of obtaining a standard medical waste multi-dimensional data stream includes: Diverting and identifying the medical waste multidimensional data stream according to the IoT sensing device group to obtain a multi-source medical waste data stream; determining a multi-source medical waste data pre-processing procedure based on characteristic information of the multi-source medical waste data stream; Preprocessing the multi-source medical waste data stream based on the multi-source medical waste data preprocessing program to obtain a usable multi-source medical waste data stream; Anomaly identification and data cleaning are performed on the available multi-source medical waste data stream to obtain a standard medical waste multi-dimensional data stream.

3. The intelligent dispatching method of medical waste robot based on Internet of Things according to claim 2, characterized in that: The method of obtaining a standard medical waste multi-dimensional data stream includes: Performing abnormality threshold analysis on the available multi-source medical waste data streams respectively to determine an abnormality judgment threshold for the multi-source medical waste data; Performing abnormal data judgment on the available multi-source medical waste data stream based on the multi-source medical waste data abnormality judgment threshold to obtain an abnormal multi-source medical waste data set; The abnormal multi-source medical waste data set is cleaned to obtain a standard medical waste multi-dimensional data stream.

4. The intelligent dispatching method of medical waste robot based on Internet of Things according to claim 1, characterized in that: The obtaining of medical waste production forecast information includes: Collecting and obtaining a historical medical waste dataset, dividing and marking the historical medical waste dataset into time series, and obtaining a historical medical waste sequence dataset; Using a time series neural network to perform model training on the historical medical waste sequence data set to generate a medical waste prediction model; Based on the medical waste prediction model, waste production prediction is performed on the standard medical waste multidimensional data stream to obtain the medical waste production prediction information.

5. The intelligent scheduling method of medical waste robot based on the Internet of Things according to claim 4, characterized in that: The generating of the medical waste prediction model comprises: Using a time series neural network to perform attribute identification training on the historical medical waste sequence data set to obtain a waste generation amount prediction model, a waste type distribution prediction model, and a waste generation peak period prediction model; The garbage generation amount prediction model, the garbage type distribution prediction model and the garbage generation peak period prediction model are combined to obtain a basic garbage prediction model; The basic waste prediction model is subjected to performance verification and optimization to generate the medical waste prediction model.

6. The intelligent dispatching method of medical waste robot based on Internet of Things according to claim 1, characterized in that: Determining the robot work scheduling plan includes: Defining a robot garbage disposal goal, and constructing a robot garbage disposal objective function based on the robot garbage disposal goal; The medical waste production prediction information and the working status information of the medical waste robot are used as constraints, and an ant colony algorithm is used to perform a scheduling solution analysis on the constraints to obtain a robot scheduling solution space; The robot garbage disposal objective function is used to perform global optimization in the robot scheduling solution space to determine the robot work scheduling solution.

7. The intelligent dispatching method of medical waste robot based on Internet of Things according to claim 6, characterized in that: The controlling the medical waste robot to process medical waste according to the robot work scheduling plan includes: Controlling the medical waste robot to process medical waste according to the robot work scheduling plan, and monitoring the robot's working conditions in real time; Perform obstacle recognition on the working road conditions of the robot to determine road obstacle information; The robot work scheduling plan is corrected based on the road obstacle information to obtain a robot work scheduling correction plan.

8. The intelligent dispatching system of medical waste robot based on Internet of Things is characterized by: The system for implementing the intelligent scheduling method of the medical waste robot based on the Internet of Things according to any one of claims 1 to 7 comprises: A sensing device installation module is used to install an Internet of Things sensing device group on the medical waste robot. The Internet of Things sensing device group includes a weight sensor, a type identification sensor, a GPS positioning device, an environmental sensor, and a video monitoring device; a data stream acquisition module, configured to activate the medical waste robot, collect and acquire a multidimensional data stream of medical waste in real time through the IoT sensing device group, and perform data cleaning on the multidimensional data stream of medical waste to obtain a standard multidimensional data stream of medical waste; A working information collection module is used to perform waste production prediction on the standard medical waste multi-dimensional data stream, obtain medical waste production prediction information, and simultaneously collect and obtain working status information of the medical waste robot; a scheduling scheme determination module, configured to perform scheduling scheme analysis based on the medical waste production forecast information and the working status information of the medical waste robot, and determine a robot work scheduling scheme; The full-process management and control module is used to control the medical waste robot to perform medical waste treatment and full-process recording according to the robot work scheduling plan, obtain the robot medical waste treatment data chain, and perform full-process traceability management through the robot medical waste treatment data chain.

Images