Bed toilet stool automatic control method based on multi-sensor fusion

Abstract

The invention relates to the technical field of health care equipment, and discloses a bed toilet stool automatic control method based on multi-sensor fusion. The method comprises the following steps of: performing time reference alignment, channel standardization and threshold binding processing by acquiring approaching attitude sensing, stress support sensing and excretion completion judgment sensing data to generate a stage criterion set; pollution factor quantification, rule table matching and instruction arrangement are executed to form an execution sequence instruction set; controlling an electric cover plate, bearing and positioning, spraying and washing, vacuum suction and ultraviolet radiation, and generating a disinfection execution record and a sealed container state; performing phase summarization, pollution risk score updating, reset strategy arrangement and parameter mirror image output, and constructing a nursing log entry and parameter updating mirror image structure. According to the invention, full-process automatic control and cross-session adaptive adjustment under multi-sensor data fusion are realized, and the intelligence, sanitation and traceability of a nursing toilet system are improved.

Description

Technical Field

[0001] This invention relates to the field of hygiene and nursing equipment technology, and in particular to an automatic control method for a commode bed based on multi-sensor fusion. Background Technology

[0002] In the field of healthcare equipment technology, existing solutions for assisting bedridden patients with toileting typically revolve around a bedpan, support components, a spray cleaning structure, a vacuum suction structure, and manual intervention by caregivers. These solutions suffer from limitations such as relying on a single trigger signal to determine completion of excretion, adhering to fixed rules for the timing of the electric toilet lid and support positioning mechanisms, and lacking comprehensive recording of the entire process. Existing methods often involve intervention by caregivers after detecting whether excretion has occurred, followed by the caregiver performing spray cleaning and vacuum suction, and then replacing the sealed container or resetting it. Furthermore, these devices often only focus on the single process of receiving waste. In scenarios involving long-term use by bedridden patients and continuous nursing workload, these solutions often lack quantification of contaminant status, have unclear triggering times for spray cleaning and vacuum suction commands, and struggle to reliably implement commands for electric toilet lid opening and closing, support positioning, ultraviolet irradiation, and the generation of nursing log entries. Existing technologies for acquiring proximity / posture perception data, force / support perception data, and excretion completion judgment perception data to form a stage criterion set, and then obtaining the nursing session ID and stage status through session number generation, stage state initialization, and nursing session binding processing, followed by contamination factor quantification, time slice aggregation, and rule table matching to generate an initial contamination risk score and drive the execution sequence instruction set, generally lack an integrated closed loop between time base alignment, channel number standardization, resource occupancy registration, instruction completion summary, resource consumption registration, and nursing log entry generation. This makes it difficult to form a consistent process from the stage criterion set to the nursing session ID, from the execution sequence instruction set to the disinfection execution record and sealed container status, and then to the parameter update mirror structure in the application scenario of automatic bedside commode control. This results in the inability of reset instructions, risk score update records, and parameter update mirror structures to be stably associated with the nursing session ID and nursing log entries in the long term. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention provides an automatic control method for a bedside toilet based on multi-sensor fusion, comprising:

[0004] Acquire proximity / attitude perception data, force / support perception data and excretion completion judgment perception data, perform time base alignment, channel number normalization and noise segment labeling, and extract proximity trend features, in-situ stability index and humidity residual curve. Perform field dictionary mapping, dimension normalization, segmentation slicing, stage threshold binding, duration statistics and abnormal segment removal processing to generate stage criterion set.

[0005] The process involves generating session numbers, initializing stage states, and binding nursing sessions to obtain nursing session IDs and stage states. It also extracts near-in-place, in-place support, and end-of-life statuses, and performs contaminant quantification, time slice aggregation, and rule table matching to generate initial contamination risk scores. Finally, it performs instruction slot arrangement, sequential constraint loading, and resource occupancy registration to generate an execution sequence instruction set.

[0006] The system executes commands to open and close the electric cover, support and position, and subscribe to in-situ monitoring; it also executes commands to spray and rinse, vacuum suction, and isolate and control the entrance of sealed containers; it executes commands to irradiate ultraviolet light, determines secondary process triggers, and assesses container capacity; and generates disinfection execution records and the status of sealed containers.

[0007] The system summarizes the timestamps of each execution phase, summarizes the completion rate of instructions, and registers resource consumption. It extracts the time taken for each phase, the dosage of cleaning agents, and the duration of ultraviolet irradiation, and performs pollution risk score updates, reset strategy arrangement, and parameter difference analysis. It generates reset instructions for the electric cover mechanism and the support positioning mechanism, as well as risk score update records associated with the nursing session ID. It performs cover and support reset control, nursing log entry generation, and parameter update mirror output operations, and constructs a mirror structure of nursing log entries and parameter updates.

[0008] Furthermore, the pollution risk assessment process also includes:

[0009] The pollution risk score update is based on the fields of stage time, cleaning agent dosage and ultraviolet irradiation duration. It is compared and updated with the initial value of the previous pollution risk score, and adjusted in stages by combining downgrade execution flag, attention item flag and container replacement prompt event flag, to generate a pollution risk score update record associated with the nursing session ID.

[0010] Furthermore, the reset strategy orchestration includes:

[0011] The reset strategy is programmed by the main control unit to generate reset commands based on the tags of items requiring attention and the container replacement prompt event tags. The reset commands indicate the reset, replacement, reset positioning, or resource replenishment status of the electric cover mechanism, the support and positioning mechanism, the spray assembly, the vacuum suction assembly, and the isolation structure of the sealed container inlet.

[0012] Furthermore, the parameter difference analysis process includes:

[0013] The parameter difference analysis process compares the stage time, cleaning agent dosage, UV irradiation duration, and downgrade execution event count with the corresponding fields of the previous nursing session, identifies the offset magnitude, generates difference fragments, and writes the difference fragments into the pollution risk score update record.

[0014] Furthermore, the summary of execution phase timestamps includes:

[0015] The phase timestamp summary includes reading the timestamps of spraying, suctioning, entrance isolation, and ultraviolet irradiation and establishing a unified timeline index to form a phase schedule, which is used for subsequent phase time reconstruction and nursing log entry generation.

[0016] Furthermore, the summary of instruction completion includes:

[0017] The instruction completion rate summary is performed within the main control unit by registering the completion flag, degradation flag, or exception event cache flag for each instruction slot in the execution sequence instruction set, forming completion rate mapping entries and recording items that need attention.

[0018] Furthermore, resource consumption registration includes:

[0019] Resource consumption registration cumulatively records the dosage of spray cleaning agent, the operating time of the negative pressure pump, the duration of ultraviolet irradiation, and the duration of support drive occupation, and associates them with the estimated available capacity of the container in the sealed container state and the replacement reminder event marker.

[0020] Furthermore, the process of updating the risk score record associated with the nursing session ID also includes:

[0021] Nursing log entries are generated by combining nursing session ID, phase schedule, instruction completion mapping entries, resource consumption registration results, contamination risk score update records, container replacement prompt event markers, and cover and support reset status.

[0022] Furthermore, the process of constructing a mirror structure for nursing log entries and parameter updates includes:

[0023] The parameter update mirror structure encapsulates the risk score update record, difference fragments, and container replacement prompt event markers, including field dictionary update information for the next nursing session, spray metering valve throttling configuration, UV irradiation cycle configuration, and container capacity determination threshold.

[0024] Furthermore, the process of constructing a mirror structure for nursing log entries and parameter updates also includes:

[0025] The parameter update mirror structure resolves the baseline and threshold baseline at the start of the next nursing session, which is used for proximity / posture perception data parsing, nursing session binding, and dynamic adjustment of contamination risk, realizing cross-session adaptive control closed loop.

[0026] The key innovations of this invention include:

[0027] (1) In the automatic control process of the bedside toilet, the approach / posture perception data, force / support perception data and excretion completion judgment perception data are acquired and time reference alignment, channel number standardization and noise segment labeling are performed. The approach trend features, in-situ stability index and humidity residual curve are extracted. After field dictionary mapping, dimension normalization, segmentation slicing, stage threshold binding, duration statistics and abnormal segment removal processing, a stage criterion set is generated. The nursing session ID and stage status are obtained through session number generation, stage status initialization and nursing session binding processing. Then, the approach position status, in-situ support status and end judgment status are extracted. Combined with pollution factor quantification, time slice aggregation and rule table matching processing, the initial value of pollution risk score is generated. The execution sequence instruction set is generated by instruction slot arrangement, sequential constraint loading and resource occupation registration processing. The stage association from stage criterion set to nursing session ID and then to execution sequence instruction set is realized.

[0028] (2) During the nursing session, based on the execution sequence instruction set, the electric cover opening and closing instruction, the support positioning instruction, and the in-situ monitoring subscription are executed. In the same nursing session, the spray rinsing instruction, the vacuum suction instruction, and the sealed container inlet isolation control are executed in sequence. Then, the ultraviolet irradiation instruction, the secondary process trigger judgment, and the container capacity assessment are executed to generate the disinfection execution record and the sealed container status. This makes the electric cover mechanism, the support positioning mechanism, the spray rinsing, the vacuum suction, the sealed container inlet isolation control, and the ultraviolet irradiation form a controlled sequence link, rather than a single trigger action.

[0029] (3) At the end of the nursing session, the execution phase timestamp summary, instruction completion summary and resource consumption registration are performed. The phase time, cleaning agent dosage and ultraviolet irradiation duration are extracted and the pollution risk score is updated, the reset strategy is arranged and the parameter difference analysis is performed. The reset instructions pointing to the electric cover mechanism and the support positioning mechanism and the risk score update record associated with the nursing session ID are generated. The cover and support reset control, nursing log entry generation and parameter update mirror output operation are further performed to construct the nursing log entry and parameter update mirror structure, so that the pollution risk score update record, reset instruction, nursing log entry and parameter update mirror structure form a closed loop, and serve as the reusable parameter basis for subsequent nursing sessions.

[0030] The following are its main beneficial effects:

[0031] (1) By continuously generating the stage criterion set, nursing session ID, nursing session ID and stage status, initial value of contamination risk score and execution sequence instruction set, the existing control process that relies on the experience judgment of nursing staff and single excretion trigger is replaced by the sequential management of the near-in-place status, in-place support status and end judgment status that are distinguished by the proximity trend characteristics, in-place stability index and humidity residual curve. In the bed-toilet scenario, the sequentially controlled action set can be accurately issued for the same nursing session, instead of manual intervention after excretion is completed. This solves the problems of lagging bed-toilet triggering logic, unclear stage division and unsystematic action instructions in the background technology.

[0032] (2) By issuing electric cover opening and closing commands, support and positioning commands, and in-situ monitoring subscriptions, and issuing spray rinsing commands, vacuum suction commands, and sealing container inlet isolation control commands in sequence within the same nursing session, combined with UV irradiation commands, secondary process trigger judgments, and container capacity assessments, disinfection execution records and sealing container status are generated. This transforms the electric cover mechanism, support and positioning mechanism, spray rinsing, vacuum suction, sealing container inlet isolation control, and UV irradiation from dispersed device behaviors into a traceable collaborative execution process. This allows for the simultaneous acquisition of disinfection execution records and sealing container status during continuous nursing care for bedridden toilet assistance, avoiding the traditional practice of only recording whether excretion has been completed and having nursing staff manually clean and replace containers, which is prone to contamination spread and lack of traceability.

[0033] (3) By summarizing the stage timestamps, summarizing the instruction completion rate, and registering the resource consumption, the stage time, cleaning agent dosage, and ultraviolet irradiation duration are associated with the nursing session ID. Further, the pollution risk score update, reset strategy arrangement, and parameter difference analysis are performed to obtain the risk score update record associated with the nursing session ID and the reset instruction pointing to the electric cover mechanism and the support positioning mechanism. At the same time, nursing log entries are generated based on the cover and support reset control and the parameter update mirror structure is output. This allows the nursing log entries and the parameter update mirror structure to be implemented in the same nursing session. They can be directly used as inputs for field dictionary mapping, dimension normalization, stage state initialization, rule table matching processing, and sequential constraint loading in the next nursing session. This avoids the situation in the background technology where the control process of the bedside toilet only focuses on a single excretion action, does not record changes in pollution factors, and does not output nursing log entries. This improves the continuous and verifiable capability between multiple nursing sessions. Attached Figure Description

[0034] Figure 1 This is a flowchart illustrating an automatic control method for a bedside toilet based on multi-sensor fusion, provided as an embodiment of this application. Detailed Implementation

[0035] Example 1: Refer to Figure 1 This is a flowchart illustrating an automatic control method for a bedside toilet based on multi-sensor fusion, provided by an embodiment of the present invention. The process may include at least steps S100-S400:

[0036] S100: Acquire proximity / attitude perception data, force / support perception data and excretion completion judgment perception data, perform time base alignment, channel number normalization and noise segment labeling, and extract proximity trend features, in-situ stability index and humidity residual curve. Perform field dictionary mapping, dimension normalization, segmentation slicing, stage threshold binding, duration statistics and abnormal segment removal processing to generate stage criterion set.

[0037] S200: Execute session number generation, stage state initialization and nursing session binding processing to obtain nursing session ID and stage state, extract near-in-place state, in-place support state and end judgment state, and perform contamination factor quantification, time slice aggregation and rule table matching processing to generate initial value of contamination risk score, and perform instruction slot arrangement, sequential constraint loading and resource occupation registration processing to generate execution sequence instruction set;

[0038] S300 executes commands to open and close the electric cover, support and position, and subscribe to in-situ monitoring; executes commands to spray and rinse, vacuum suction, and isolate and control the entrance of the sealed container; executes commands to irradiate ultraviolet light, determines the secondary process trigger, and assesses the container capacity; and generates disinfection execution records and the status of the sealed container.

[0039] S400, summarizing execution phase timestamps, instruction completion rates, and resource consumption, extracting phase time, cleaning agent dosage, and UV irradiation duration, and performing contamination risk score updates, reset strategy arrangement, and parameter difference analysis, generating reset instructions for the electric cover mechanism and support positioning mechanism, as well as risk score update records associated with the nursing session ID, executing cover and support reset control, nursing log entry generation, and parameter update mirror output operations, and constructing a nursing log entry and parameter update mirror structure.

[0040] S100: Acquire proximity / attitude perception data, force / support perception data, and excretion completion judgment perception data; perform time base alignment, channel number normalization, and noise segment labeling processing; extract proximity trend features, in-situ stability indicators, and humidity residual curves; perform field dictionary mapping, dimension normalization, segmentation slicing, stage threshold binding, duration statistics, and abnormal segment removal processing; and generate a stage criterion set.

[0041] Specifically, the S100 process first uses the sensor acquisition side of the automatic control device for the commode bed as the input source. The sensor acquisition side includes a proximity / posture sensing unit for acquiring proximity / posture sensing data, a force / support sensing unit for acquiring force / support sensing data, and a defecation completion judgment sensing unit for acquiring defecation completion judgment data. The proximity / posture sensing unit includes an infrared ranging sensor and a human body pyroelectric sensor arranged in the area above and around the commode bed. The infrared ranging sensor is used to acquire the distance evolution information between the patient's lower body and the commode bed support opening, and the human body pyroelectric sensor is used to acquire the thermal radiation profile and proximity direction information of the patient's pelvic region and groin region. The pressure / support sensing unit includes a zoned pressure sensing element, which is attached to different support areas of the bedpan support ring, support pad, or auxiliary support structure to record the pressure distribution, duration of continuous pressure, and stability of the patient's buttocks and pelvis in the support area. The excretion completion determination sensing unit includes a contamination-resistant humidity detection element, a liquid residue detection element, and a local surface humidity tracking element located in the area near the drainage port and the area near the spray nozzle. This part of the element collects the humidity status, liquid residue status, and humidity decay trend of the inner wall of the bedpan, the drainage port inlet, and the area near the spray cleaning area to characterize whether excretion has occurred and whether it has been basically completed in subsequent processing. Furthermore, the data acquired from the aforementioned sensor acquisition side is not directly analyzed after acquisition. Instead, it is input to a microcontroller unit (MCU) or a host control board that works in conjunction with the MCU. The MCU loads the acquisition strategy, trigger threshold, sampling rhythm, and other operating parameters into the parameter update mirror structure generated from the previous nursing session, thus creating a loop relationship between the input side in this step and the parameter update mirror structure generated in S400. In this input stage, proximity / attitude perception data, force / support perception data, and excretion completion determination perception data are uniformly registered as a raw signal sequence to be processed. Each sensor channel is bound with a unique channel number, physical installation area description, acquisition timestamp, and sensor health flag, forming a raw sampling set with time and source tags.

[0042] Specifically, after obtaining the original sample set, the S100 process performs time reference alignment, channel number normalization, and noise segment labeling. Time reference alignment refers to rearranging the sampling events of different types of sensors according to a unified time sequence, making the approach / posture behavior, support force behavior, and residual humidity behavior comparable within the same time slice. In this alignment process, if a sensor channel has a sampling interval that significantly lags behind other channels, that channel is marked as an intermittent segment in that interval and will not participate in subsequent stability calculations. Channel number normalization refers to establishing a mapping relationship between the aforementioned unique channel number and the fixed physical index of the bedpan body, so that the approach / posture sensing units, force / support sensing units, and excretion completion judgment sensing units in the same physical area can be aggregated by region in subsequent processing stages. This channel number normalization also records the calibration version and installation version of the sensor channel, facilitating the identification of structural displacement or replacement behavior within the same nursing session. Noise segment labeling refers to marking abnormal high-amplitude impact segments, transient jump segments, and corresponding channel self-testing abnormal sampling segments in the original sampling set. Transient jump segments include short-term high-pressure impacts caused by the patient being moved by others, transient meaningless close-range triggers caused by bed shaking, and short-term high-humidity spikes caused by residual water mist from showering. These segments are removed in subsequent feature extraction stages and are not included in stability or excretion completion determination. After the above processing, a time-aligned, channel-unified, and noise-labeled original perception data package is obtained. This original perception data package is used as the starting point for state feature extraction in subsequent processing.

[0043] Furthermore, after obtaining the raw sensing data, the S100 process performs a state feature extraction process, which includes proximity trend feature extraction, in-situ stability index extraction, and humidity residual curve extraction. Proximity trend features refer to a set of features reflecting whether the patient's lower body or pelvic region is continuously approaching the bedpan support opening. Specifically, the system reads the distance evolution information from the infrared ranging sensor from the proximity / posture sensing data and combines it with the thermal distribution profile stability of the human body pyroelectric sensor to identify in time whether the approach direction, approach speed, and approach duration occur continuously, rather than a single instantaneous approach followed by immediate departure; this part of the information is uniformly abstracted as proximity trend features. In-situ stability indicators refer to a set of indicators used to describe whether a patient has been stably positioned and maintained in a supported state. Specifically, the system reads the regional pressure distribution and duration of pressure on the zoned pressure sensing pads from the force / support sensing data, and determines whether the pressure distribution has entered a relatively fixed state. If it maintains a similar pressure profile for a long time without significant load jumps, the segment is marked as a candidate segment for in-situ stability. The pressure balance and pressure holding time of each zone in the candidate segment are summarized into in-situ stability indicators. The residual humidity curve refers to a type of curve information used to describe whether excreted liquids or wet excrement have appeared and are gradually decaying. Specifically, the system reads the evolution of the humidity and residual amount states of the humidity detection element and the liquid residue detection element over time from the excretion completion judgment sensing data, and slices these evolution processes over time, arranging the obviously high humidity rising segment, plateau segment, and falling segment in chronological order to form the residual humidity curve. The aforementioned proximity trend characteristics, in-situ stability indicators, and residual humidity curves together constitute a state feature vector.

[0044] After obtaining the state feature vector, the S100 process continues with field dictionary mapping, dimension normalization, segmentation, stage threshold binding, duration statistics, and abnormal segment removal. Field dictionary mapping maps the descriptive fields from different sensor channels to a unified field dictionary. This field dictionary, maintained by a field dictionary mapping table in the device firmware, converts descriptions of human approach behavior from infrared ranging sensors, force balance from partitioned pressure sensors, and humidity decay from humidity detection elements into unified semantic fields for easy subsequent use. Dimension normalization standardizes the numerical ranges of different physical measurements, allowing for threshold comparisons and duration statistics of approach trend features, in-situ stability indicators, and residual humidity curves on the same scale. Segmentation divides the state feature vector into several continuous time segments on a unified time axis, with each segment corresponding to a candidate window for the approach, in-situ, or end stage. Stage threshold binding refers to matching the pre-stored threshold configuration within the device with the feature intensity of the aforementioned time segment to generate an initial stage label. The stage label indicates that the segment is closer to the in-situ or end stage. This threshold configuration can be revised by the parameter update mirror structure of the previous nursing session, thereby applying the parameter update mirror structure generated in S400 to the criterion generation in this step. Duration statistics refers to calculating the duration span of the time segments with initially bound stage labels, used to distinguish between transient interference events and stable state events. Abnormal segment removal refers to removing transient interference segments discovered in the duration statistics, as well as segments previously marked with noise segment annotations, from the candidate stage events and preventing them from entering subsequent nursing session generation stages.

[0045] Finally, after field dictionary mapping, dimension normalization, segmentation, stage threshold binding, duration statistics, and outlier removal, a stage criterion set is generated. The stage criterion set consists of several time-aligned stage event entries. Each stage event entry records effective segments of three types of information: proximity trend characteristics, in-situ stability indicators, and residual humidity curves, along with the corresponding stage label and the start and end times of that segment. The stage criterion set is directly passed as an output field to subsequent steps. Specifically, it is read in S200 during session number generation, stage state initialization, and nursing session binding to generate the nursing session ID and stage state. The nursing session ID and stage state then participate in contaminant quantification, rule table matching, and instruction slot arrangement to form an execution sequence instruction set. In S300, this sets drive the issuance of commands for electric cover opening / closing, support positioning, spray rinsing, vacuum suction, sealed container inlet isolation control, ultraviolet irradiation, and secondary process triggering. The disinfection execution records and sealed container states of the aforementioned actions are collected and summarized in S400 to generate a process summary package, reset commands, and risk score update records. Finally, the output parameter update mirror structure is fed back to this step as the configuration baseline and criterion baseline for the proximity / posture sensing unit, force / support sensing unit, and excretion completion judgment sensing unit, thus creating a closed-loop relationship between the generation of the stage criterion set and the subsequent nursing session state generation.

[0046] Therefore, the technical effect of this step can be summarized as follows: by uniformly accessing, standardizing and segmenting the proximity / posture perception data, force / support perception data and excretion completion judgment perception data, a set of staged criteria can be formed that can directly drive the generation and execution sequence of subsequent nursing sessions. In the overall process of the automatic bedpan control method, a cross-step closed loop of acquisition, judgment, execution and reset is established.

[0047] S200: Execute session number generation, stage state initialization and nursing session binding processing to obtain nursing session ID and stage state, extract near-in-place state, in-place support state and end judgment state, and perform contamination factor quantification, time slice aggregation and rule table matching processing to generate initial value of contamination risk score, and perform instruction slot arrangement, sequential constraint loading and resource occupation registration processing to generate execution sequence instruction set.

[0048] Specifically, S200 takes a stage criterion set as input. This stage criterion set is a set of staged events obtained in the preceding step S100 through time base alignment of proximity / posture sensing data, force / support sensing data, and excretion completion judgment sensing data; channel number normalization; noise segment labeling; field dictionary mapping; dimension normalization; segmentation; stage threshold binding; duration statistics; and abnormal segment removal. Each stage event entry in the staged event set includes proximity trend characteristics, in-situ stability indicators, humidity residual curves, stage labels, and their start and end time information. S200 is executed by the main control unit of the automatic bedpan control device. The main control unit can be a microcontroller unit (MCU) or a control processing unit composed of a microcontroller unit and a host control board. After receiving the stage criterion set, the main control unit first performs session number generation processing. This process generates a unique nursing session number for each newly appearing stage event entry in the current stage criterion set. This unique nursing session number is referred to as the nursing session ID in this description. The nursing session ID is used to bind a complete bedridden toileting process with the corresponding sensor data segment, execution command segment, and disinfection reset segment. Once generated, the nursing session ID remains valid throughout the entire nursing process and will not be repeatedly assigned during the same process. Subsequently, the master control unit performs stage state initialization processing. Stage state initialization processing refers to marking the initial nursing stage of the current nursing session as one of three candidate states: approaching preparation, in-situ support, or end confirmation, based on the stage labels in the stage criterion set. During initialization, the master control unit registers the nursing session ID along with the initial nursing stage, forming a structured binding record of nursing session ID and stage state. The nursing session ID and stage state are the first type of output field in this section, and this output field will be continuously referenced as input in subsequent steps.

[0049] Furthermore, after obtaining the nursing session ID and stage status, the main control unit performs extraction processing for the approach positioning state, in-place support state, and termination judgment state for different stages of the nursing process. The approach positioning state refers to the state in which the patient's body parts (especially the pelvic region and the groin region) are continuously close to the bedpan support opening and form a stable approach posture in physical space. This state corresponds to the stage event item in the approach trend characteristics that shows continuous approach without rapid departure. The in-place support state refers to the state in which the patient's buttocks and pelvic region are stably supported by the support ring, support pad, or other support structure, and the force distribution in the zoned pressure sensing pad shows a relatively stable force profile and continuous pressure holding time. This state corresponds to the stage event item marked as stable holding in the in-place stability index. The termination judgment state refers to the state in which the excretion behavior has occurred and tends to end, and the residual humidity curve shows a trend of transitioning from the high humidity plateau segment to the downward segment, while no new high humidity rising segment appears in the subsequent time slice. This state corresponds to the stage event item marked as the termination stage in the residual humidity curve. The main control unit extracts these three types of states from the nursing session ID and stage status. The extraction process uses the nursing session ID as an index, aligning the start and end times of the near-in-place state, the in-place support state, and the end-of-term determination state on the timeline. It also associates the temporal boundaries of these three states with their respective nursing session IDs, ensuring that each nursing session ID has traceable markers for the near-in-place period, the in-place period, and the end-of-term period. Understandably, the extraction of the near-in-place state, the in-place support state, and the end-of-term determination state here not only transforms stage labels into actionable stage states but also sequentially binds these stage states with the nursing session ID, providing structured state clues for subsequent steps.

[0050] After completing the above state extraction, the main control unit performs contamination factor quantification, time slice aggregation, and rule table matching on the current nursing session to generate an initial contamination risk score. Contamination factor quantification refers to the quantitative registration of key contamination-related elements appearing in the in-situ support state and the end-of-life judgment state corresponding to the nursing session. Key contamination-related elements include at least the duration of the high-humidity plateau segment corresponding to the end-of-life judgment state in the residual humidity curve, the rate of decrease in the humidity decay segment, the expected scale of residual discharge from the area to be treated before aspiration, and abnormal fluctuations in local pressure in the support area during the in-situ support state. These key contamination-related elements are matched with the residual humidity curve and in-situ stability indicators in the stage criterion set to form a set of contamination factor entries. Time slice aggregation refers to the main control unit splicing and sorting the time slices corresponding to the three states—approaching in place, in-situ support, and end-of-life judgment—on the time axis according to the nursing session ID, resulting in a nursing time sequence arranged in the actual order of occurrence. This nursing time sequence is used to describe the temporal organization of the complete nursing process from approaching to in-situ to end. The rule table matching process refers to the main control unit comparing the set of contaminant entries and the sequence of nursing time segments with a pre-stored rule table. The rule table records the stage intensity definition information at the time of device manufacturing and the reset strategy arrangement results carried in the parameter update mirror structure output by S400 in the previous nursing cycle. The rule table maps different combinations of contaminants and different durations to contamination grading segments. After completing this comparison, the main control unit generates an initial contamination risk score for the nursing session. In this description, the initial contamination risk score refers to the qualitative number and segment description of the contamination grading segment of the current nursing session. This qualitative number and segment description do not directly control hardware actions but will be used to indicate the execution intensity in subsequent instruction slot arrangement. This initial contamination risk score is the second type of output field, which will be called by subsequent instruction slot arrangement, sequence constraint loading, and resource occupancy registration processing.

[0051] Subsequently, the main control unit performs instruction slot orchestration, sequence constraint loading, and resource occupancy registration to generate an execution sequence instruction set. Instruction slot orchestration refers to establishing an execution sequence consisting of multiple types of actions for this nursing session. These actions include electric cover opening and closing instructions, electric support structure support and positioning instructions, in-situ monitoring subscription during the in-situ process, spray rinsing instructions during the cleaning phase, vacuum suction instructions during the excrement removal phase, sealed container inlet isolation control when excrement enters the sealed container, and ultraviolet irradiation instructions and secondary process trigger determination during the terminal hygiene treatment phase. Each action is assigned to an independent instruction slot, which is a issuing unit within the main control unit used to specify the specific execution component, triggering time, allowed execution duration, and available resources. Sequential constraint loading refers to arranging these instruction slots sequentially according to the state order of the nursing session ID: During the period corresponding to the near-in-position state, electric cover opening / closing instructions and support / positioning instructions are inserted; during the period corresponding to the in-position support state, in-position monitoring subscription is inserted; during the period corresponding to the end judgment state, spray rinsing instructions, vacuum suction instructions, sealed container inlet isolation control, ultraviolet irradiation instructions, and secondary process trigger judgments are inserted. The result of sequential constraint loading is an execution sequence chain that cannot be arbitrarily rearranged. As the stage state of the nursing session ID is advanced, the main control unit will release the corresponding instruction slots one by one according to this sequence chain. Resource occupancy registration processing refers to registering the occupancy of hardware resources involved in the above-mentioned execution actions. Hardware resources include electric cover drive mechanisms, support / positioning components, spray nozzle components, vacuum suction components, sealed containers and their one-way isolation structures, ultraviolet irradiation components, etc. Resource occupancy registration processing records the locking status and borrowing period of these hardware resources under the current nursing session ID to prevent parallel sessions from contending for the same spray nozzle or the same vacuum suction channel, leading to execution conflicts. Through instruction slot arrangement, sequential constraint loading, and resource occupancy registration, the main control unit forms a complete execution sequence instruction set. This set includes action descriptions, trigger relationships, and resource locking relationships for all execution actions arranged in chronological order. The execution sequence instruction set is the third type of output field in this section. This output field is directly acquired and issued in subsequent step S300. After receiving the execution sequence instruction set, S300 sequentially issues the electric cover opening / closing instruction, the support positioning instruction, and the in-situ monitoring subscription; executes the spray rinsing instruction, the vacuum suction instruction, and the sealed container inlet isolation control; and executes the ultraviolet irradiation instruction, the secondary process trigger judgment, and the container capacity assessment, generating a disinfection execution record and the sealed container status.Therefore, the output of S200 is used as the input of S300, and the disinfection execution record and sealed container status generated by S300 are read by S400 and written into the process summary package and reset instructions and risk score update records. The reset instructions and risk score update records are finally output by S400 as nursing log entries and parameter update mirror structure, and are then fed back to S100 and this step in the next nursing session for use, thus forming a closed loop across S100, S200, S300 and S400.

[0052] The technical effects of this step can be summarized as follows: By generating session numbers, initializing stage states, and binding nursing sessions to the stage criterion set, nursing session IDs and stage states are obtained. Based on this, the near-in-place state, in-place support state, and end judgment state are uniformly extracted. The main control unit can quantify contaminants, aggregate time slices, and match rule tables to output an initial value of the contamination risk score. Subsequently, the initial value of the contamination risk score is used for instruction slot arrangement, sequential constraint loading, and resource occupation registration to generate an execution sequence instruction set and pass it to the S300 for execution. The stage criterion set obtained by multiple sensors is directly converted into an executable nursing session drive sequence.

[0053] S300 executes commands to open and close the electric cover, support and positioning, and subscribe to in-situ monitoring; executes commands to spray and rinse, vacuum suction, and isolate the sealed container inlet; executes commands to irradiate ultraviolet light, determines secondary process triggers, and assesses container capacity; and generates disinfection execution records and sealed container status.

[0054] Specifically, this step takes the execution sequence instruction set as input, which is output from the previous step S200. The execution sequence instruction set has already established a timing link and resource locking relationship according to the nursing session ID and stage status, and also includes an initial contamination risk score to indicate the intensity of subsequent execution. In this step, the microcontroller unit (MCU) loads the instruction slot queue of the current nursing session and releases the electric cover opening / closing command, support positioning command, and on-site monitoring subscription in the order of near-in position and in-situ support status. The electric cover mechanism includes a motor drive assembly, a mechanical travel sensor, and limit switches. After receiving the corresponding command, the MCU first issues unlock and slow-open commands, reads the displacement of the mechanical travel sensor and the contact status of the limit switches. If the contact status is within the allowable window and the displacement continuously increases, the cover is determined to enter the opening process. If displacement stagnation or abnormal contact status occurs within the preset monitoring window, the cover opening / closing abnormal event is marked in the local event cache of the current nursing session, and the timestamp and channel number of the abnormal event are recorded. The support and positioning mechanism includes a support lifting component and a support angle adjustment component. During the support and positioning command issuance phase, the MCU issues lifting and angle adjustment actions in two stages based on the pressure zone target in the in-situ support state. During this period, it reads the support stroke sensor and the support drive current changes to identify whether there is jamming or load abnormality. If three consecutive micro-displacement failures are detected, the command slot is marked as degraded execution and written to the event cache. In-situ monitoring subscription is an asynchronous event subscription to the zoned pressure sensing chip and the proximity / attitude sensing unit. The MCU registers a callback channel to receive continuous segments of in-situ stability indicators and reverse fluctuation information of proximity trend characteristics to determine whether to pause subsequent spraying and suction in the current session. When the subscription event shows that the in-situ stability indicator has entered the stable range and no rapid departure trend has appeared, the MCU writes the in-situ monitoring subscription ready into the execution pipeline record and enters the next processing stage.

[0055] Furthermore, during the spray flushing command issuance, vacuum suction command issuance, and sealed container inlet isolation control processing stages, the MCU reads the action sequence and resource occupancy registration from the execution sequence instruction set, and issues synchronization commands to the spray assembly and fluid control assembly via the internal control bus. The spray assembly includes spray nozzles, detergent metering valves, and a water pump. The MCU first sets the number of metering valve opening pulses and the spray duration based on the initial value of the contamination risk score, then starts the water pump to establish the spray pressure, and continuously reads the flow pulsation signal near the nozzle during spraying to identify whether the nozzle is blocked; if an abnormal flow pulsation amplitude occurs within the monitoring window, a short pause in spraying is triggered and the nozzle self-cleaning subroutine is executed. Fluid suction is accomplished by a vacuum suction assembly, which includes a negative pressure pump, a suction channel pressure sensor, and a condensation recovery unit. After the spray system starts, the MCU enters the vacuum suction phase after a delay, reading the suction channel pressure curve and the changes in the negative pressure pump's operating current in real time to determine whether suction is continuous and whether the channel is unblocked. When the pressure curve fluctuates beyond limits or the current continues to rise, the MCU executes an intermittent suction and backflow mitigation strategy and registers the suction anomaly event in the event cache. The sealed container inlet isolation control consists of a one-way isolation structure at the container inlet and an electric gate. After suction stabilizes, the MCU issues an opening command and reads the opening feedback from the inlet position sensor. Simultaneously, it compares the liquid level response from the liquid level sensor inside the container. If the opening feedback is normal and the liquid level response increment is synchronized with the suction rate, the gate remains open; otherwise, the MCU closes the gate and enters a short-term backflow to prevent external backflow into the suction channel. The coordinated actions of spraying, suctioning, and entrance isolation are all recorded according to the nursing session ID, including time segments and instruction completion status. During this process, three types of sub-records are generated: spraying dose record, suction duration record, and entrance isolation status record. These are temporarily stored in the local buffer to provide the original basis for the subsequent generation of disinfection execution records.

[0056] The process then proceeds to the stages of issuing UV irradiation commands, determining secondary process triggers, and assessing container capacity. The UV irradiation assembly consists of UV lamps, a reflector cavity, and a door / cover closure detection sensor. After confirming that the cover is in an irradiable position and the door / cover closure detection sensor returns a closure indication, the MCU issues a UV irradiation command, sets the irradiation cycle length according to the initial pollution risk score, and monitors the UV lamp's illumination current and temperature rise curve to determine if the irradiation is stable. If abnormal fluctuations in the illumination current occur within the preset irradiation period, the MCU performs irradiation period compensation and cooling intervals. The secondary process trigger determination is a combined determination of the spray dosage record, aspiration duration record, and humidity residual curve readback results. The MCU compares the residual decrease of the local humidity probe before and after irradiation with the matching relationship between the spray and aspiration input levels. When the matching relationship is outside the allowable range, a short process of secondary spraying and secondary aspiration is initiated based on the initial pollution risk score, and the triggering reason, triggering time, and secondary execution dosage are included in the execution log record of this nursing session. The container capacity assessment is based on the liquid level sensor and weight recognition unit. After the suction and irradiation phases, the MCU compares the current liquid level and weight increment of the container to generate an estimated value of the available remaining capacity. Based on the estimated available remaining capacity and the historical suction rate of the session, it determines whether the container is approaching its capacity limit. When it is close to the limit, the MCU registers a container replacement prompt event in its local event cache for subsequent unified aggregation and generation of reminder entries by the S400. Through the above processing, the MCU aggregates various fields of spraying, suction, inlet isolation, and ultraviolet irradiation according to the nursing session ID, combines abnormal event cache and secondary process trigger information, and archives them to form a disinfection execution record for this session. At the same time, it reads the container liquid level, weight, and inlet gate position status to generate the corresponding sealed container status.

[0057] Understandably, the output of this step uses the disinfection execution record and sealed container status as output field names, and is written to a persistent storage queue at the end of the step for direct access by the disinfection execution record and sealed container status input location in subsequent step S400. After reading this output, S400 performs stage timestamp summarization, instruction completion rate summarization, and resource consumption registration, and accordingly completes the contamination risk score update and reset strategy arrangement. The resulting parameter update mirror structure is fed back to S100 and S200 in the next nursing session to revise the field dictionary mapping and the strength configuration of the execution sequence instruction set, thereby forming a closed loop of cross-step data and actions. The technical effect of this step can be summarized as follows: by segmenting the execution sequence instruction set and driving multiple components collaboratively, the sequential execution of cover plate, support, spray, suction, isolation, and ultraviolet irradiation is completed, and the disinfection execution record and sealed container status that can be statistically analyzed and traced are generated using the nursing session as a carrier, providing directly callable execution surface data for subsequent reset and parameter updates.

[0058] S400, summarizing execution phase timestamps, instruction completion rates, and resource consumption, extracting phase time, cleaning agent dosage, and UV irradiation duration, and performing contamination risk score updates, reset strategy arrangement, and parameter difference analysis, generating reset instructions and risk score update records, executing cover and support reset control, nursing log entry generation, and parameter update mirror output operations, and constructing a nursing log entry and parameter update mirror structure.

[0059] Specifically, this step uses the disinfection execution record and the sealed container status as inputs. The disinfection execution record and the sealed container status are generated by the microcontroller unit (MCU) for the same nursing session ID in the preceding step S300. The disinfection execution record describes the execution process and sequence of the spray flushing command, vacuum suction command, sealed container inlet isolation control, ultraviolet irradiation command, and secondary process trigger determination in the nursing session. It includes spray dosage record, suction duration record, inlet isolation status record, ultraviolet irradiation segment illumination time information, secondary process triggering reason information and triggering time, etc. The sealed container status describes the available capacity of the sealed container after the end of the nursing session, inlet gate position, liquid level sensor feedback, weight recognition unit feedback and other resource status information. This step first initiates the stage timestamp aggregation within the main control unit. Stage timestamp aggregation is a process of collecting timestamps for each key execution node of the nursing session and establishing a unified timeline index. This process reads the timestamps marked as spray start, spray end, suction start, suction end, entrance isolation open, entrance isolation closed, UV irradiation start, and UV irradiation end from the disinfection execution record and stores these timestamps sequentially into the stage timetable. Next, the execution instruction completion rate is aggregated. This process compares and registers the actual execution status of each instruction slot in the execution sequence instruction set (output from S200 and issued by S300). The main control unit reads the disinfection execution records one by one. The completion identifier, degradation identifier, or abnormal event cache identifier of each executed action in the line record is mapped to the target instruction item in the execution sequence instruction set corresponding to the nursing session, forming a completion mapping entry. In the above process, if an instruction slot is found to be marked as degraded execution or has a stuck event in the S300 stage (e.g., displacement stagnation event in the electric cover opening and closing process, micro-displacement failure event in the support and positioning process, suction abnormality event in the vacuum suction process, nozzle self-cleaning subroutine trigger event in the spraying process, and lighting compensation event in the ultraviolet irradiation process), then the instruction slot is registered as an item requiring attention, and this item requiring attention will be given priority in the subsequent reset strategy arrangement. Alongside the summary of stage timestamps and instruction completion, the main control unit also performs resource consumption registration. Resource consumption registration is the process of cumulatively registering measurable resources such as spray cleaning agent, working time of suction negative pressure pump, continuous irradiation period of ultraviolet irradiation device, time occupied by support and positioning mechanism, and power-on period of electric cover drive mechanism during nursing sessions. The registration results include resource usage description fields such as cleaning agent dosage, negative pressure pump working time, ultraviolet irradiation time, and support drive time occupied, and are associated with the estimated available container balance in the sealed container status and the container replacement prompt event flag.After the process of summarizing stage timestamps, instruction completion rates, and resource consumption, the above information is organized into a process summary package. The process summary package is used as an intermediate derived structure in this step and is not output to other steps separately, but it directly participates in the analysis and processing of the next sub-process.

[0060] Further, after obtaining the process summary package, this step extracts the stage time, cleaning agent dosage, and UV irradiation duration, and performs contamination risk score updates, reset strategy arrangement, and parameter difference analysis to generate reset instructions and risk score update records. Stage time refers to the result of reconstructing the duration of the near-in-place state, in-place support state, and end-of-concept state for the same nursing session. This duration information is calculated by pairing timestamps obtained from the stage timestamp summary and archived in segments according to the nursing session ID. Cleaning agent dosage refers to the cleaning agent dosage description field accumulated through the number of metering valve pulses during the spray rinsing instruction period, which is directly extracted from the resource consumption registration. UV irradiation duration refers to the cumulative period during which the UV irradiation device remains lit during the UV irradiation instruction period, which is directly extracted from the resource consumption registration. These three fields are compared and updated with the initial pollution risk score generated in step S200 within this step. The pollution risk score update process is carried out by the main control unit based on the downgrade execution flag, the attention item flag, and the container replacement prompt event flag in the process summary package. The adjustment involves merging the original segmented description with the actual resource consumption intensity, stage time extension, spray dose distribution, and ultraviolet irradiation compensation of this nursing session, thereby generating a new pollution risk score update record. The pollution risk score update record is a record unit with an associated session ID, used to describe the final pollution treatment intensity, anomaly distribution, and resource usage characteristics of this nursing session. Simultaneously, this step involves reset strategy orchestration. Reset strategy orchestration refers to the process by which the main control unit generates subsequent reset instructions based on the attention item markers and container replacement prompt event markers. The reset instructions are used to indicate which hardware components need to enter reset replacement, reset positioning, or resource replenishment states after the nursing session ends. For example, the electric cover is restored to the closed and locked angle position, the support positioning mechanism returns to the standby support mode and releases the support drive occupation, the metering valve of the spray assembly is reset to the initial throttling position, the vacuum suction assembly stops the negative pressure pump and clears the suction channel residue, and the sealed container inlet gate is closed and locked in the isolation state. At the same time, when the container capacity is close to the boundary, the container replacement prompt event is recorded as a reference prompt for subsequent manual intervention. Parameter difference analysis and processing are carried out throughout the process of updating and resetting pollution risk scores. Parameter difference analysis and processing refers to the main control unit comparing the stage time, cleaning agent dosage, ultraviolet irradiation duration, and downgrade execution event count of the current nursing session with the corresponding fields of the previous nursing session, identifying the offset between fields of the same type, and marking the offset as a difference segment. The difference segment does not directly drive hardware execution, but is written into the risk score update record, so that it can be referenced by S100 and S200 at the beginning of the next nursing session.After the pollution risk score is updated, the reset strategy is arranged, and the parameter difference is analyzed, the main control unit generates a reset command and a risk score update record. The reset command points to the reset sequence of specific hardware components and the resource replenishment sequence. The risk score update record is used to represent the final pollution risk classification and resource usage profile of this nursing session. Both serve as direct control inputs for subsequent actions of S400.

[0061] Understandably, after generating the reset command and risk score update record as described above, this step continues to execute the cover and support reset control, nursing log entry generation, and parameter update mirror output operations, ultimately constructing the nursing log entry and parameter update mirror structure. The cover and support reset control is the execution process of the reset command. The main control unit sends commands to the electric cover mechanism and the support positioning mechanism one by one: For the electric cover mechanism, a closing action is issued, and the final displacement of the mechanical stroke sensor and the closing state of the limit switch are monitored. The final displacement of the electric cover in the closed state and the contact state of the closed limit switch are registered as the cover reset state; For the support positioning mechanism, a falling action and an angle reset action are issued, and the falling end reading of the support stroke sensor and the falling segment current characteristics of the support drive current are read. This information is registered as the support reset state. Nursing log entries are generated during the aforementioned reset process. The main control unit combines the nursing session ID, phase schedule, instruction completion mapping entries, resource consumption registration results, contamination risk score update records, container replacement prompt event markers, and cover reset status and support reset status into structured text entries, which are defined as nursing log entries. Nursing log entries are the traceable recording units for this nursing session and can be directly provided to the nursing record system or nursing handover system as nursing documentation. The parameter update mirror output operation occurs after the nursing log entries are generated. The main control unit extracts the risk score update records, difference fragments, and container replacement prompt event markers, organizes these data, and encapsulates them into a parameter update mirror structure. This parameter update mirror structure is a dataset defined in this specification for transferring control strategies and threshold configurations across nursing sessions. This dataset includes at least the field dictionary update information for the next nursing session, the dimensionless boundary revision information, the initial throttling configuration of the spray metering valve, the initial configuration of the ultraviolet irradiation cycle, the upper limit configuration of the negative pressure pump working period of the vacuum suction component, and the initial threshold for container capacity boundary determination. The nursing log entries and parameter update mirror structure are the final output fields of this step. The nursing log entries are recorded as nursing traces outside the system, while the parameter update mirror structure will be used by S100 as the parsing baseline for proximity / posture perception data, force / support perception data and excretion completion judgment perception data at the beginning of the next nursing session, and by S200 as the threshold baseline for session number generation, stage state initialization and nursing session binding processing, contaminant quantification and rule table matching processing, so that the method of the present invention has continuous adjustment capability in subsequent nursing sessions.

[0062] The technical effects of this step can be summarized as follows: By summarizing the disinfection execution records and the status of the sealed containers through phase timestamps, instruction completion rates, and resource consumption registration, the main control unit can generate a process summary package. Based on this, it can complete the update of the contamination risk score, the arrangement of reset strategies, and the analysis and processing of parameter differences, thereby obtaining reset instructions and risk score update records. During the execution of the reset instructions, the cover reset status and the support reset status are formed. Combined with the nursing session ID and resource consumption registration content, nursing log entries are generated. At the same time, the parameter update mirror structure is output and fed back to S100 and S200, thereby realizing data closed loop, instruction closed loop, and configuration closed loop.

[0063] In one embodiment, this step uses the disinfection execution record and the sealed container status as inputs. The disinfection execution record consists of the spray dosage record, aspiration duration record, entrance isolation status record, UV irradiation segment illumination time information, and secondary process triggering reason information from the previous nursing session. The sealed container status consists of the estimated available container capacity, entrance gate position, liquid level sensor feedback, and weight recognition unit feedback. The main control unit receives the above inputs at the end of the session loop, binds them with the nursing session ID of the current nursing session, and enters the stage timestamp summary process. The phase timestamp summary aligns the timestamps marked as spray start, spray end, suction start, suction end, entrance isolation open, entrance isolation closed, and UV irradiation start and end in the disinfection execution record to establish a phase timetable. Subsequently, the main control unit summarizes the completion rate of execution instructions, comparing each action's completion flag, degradation flag, and abnormal event cache flag with the execution sequence instruction set determined in S200 and issued in S300, generating completion rate mapping entries. Entries experiencing displacement stagnation, micro-displacement failure, suction abnormalities, nozzle self-cleaning subroutine triggering, and lighting compensation are marked as items requiring attention. Simultaneously, the main control unit performs resource consumption registration, accumulating measurable resources such as cleaning agent dosage, negative pressure pump operating time, UV irradiation time, support drive occupancy time, and electric cover drive power-on time. This is associated with the estimated available container capacity in the sealed container status and the container replacement prompt event flag, and archived together with the nursing session ID as a process summary package. To obtain measurable intermediate indicators, the master control unit extracts stage time, cleaning agent dosage, and UV irradiation duration from the process summary package, obtaining standardized inputs for risk updates and strategy orchestration. To unify dimensions and scales, the master control unit updates the scale configuration carried in the mirror structure based on the parameters output from the previous nursing session in S400, performs normalization mapping and weighted noise filtering, and generates the feature quantities required for risk assessment. Subsequently, to construct measurable and controllable intermediate indicators, the master control unit performs a two-step fusion calculation on the time and resource axes. Formula ①

[0064]

[0065] in:

[0066] This is the stage resource vector after scale mapping;

[0067] Update the scale matrix of the mirror structure from the parameters;

[0068] This represents the resource vector for the initial stage.

[0069] , , These represent the durations of the near-in-place phase, the in-place support phase, and the end-of-conclusion phase, respectively.

[0070] Indicates the dosage of cleaning agent;

[0071] Indicates the duration of ultraviolet irradiation;

[0072] The transpose operator represents the transpose of a row vector into a column vector or vice versa.

[0073] The mapping from data source to metric to variable is as follows:

[0074] Extract the "Spray Dosage Record" from the "Disinfection Execution Record" and record it as follows: The "UV Irradiation Period Information" is extracted from the "Disinfection Execution Record" and recorded as follows: Extract "Stage Time" from "Stage Timestamp Summary" and record them separately. , , Extract the "scale configuration" from the "parameter update mirror structure" and construct it. The output of formula ① This serves as a standardized input for subsequent risk and strategy calculations in this step, and is used in Formula ② within this section, while also being one of the input items for S420. Formula ②

[0075]

[0076] in:

[0077] The stage cost scalar after completion penalty;

[0078] This is a quadratic metric operator based on the Euclidean norm, used to measure deviation;

[0079] The stage target vector is updated from the parameter mirror structure;

[0080] To penalize the weighted scalar;

[0081] The anomaly strength scalar is obtained by aggregating completion mapping entries.

[0082] The mapping from data source to metric to variable is as follows:

[0083] Extract "Degradation Flags and Exception Event Cache Flags" from "Instruction Completion Summary" and aggregate them into The "stage goal" is extracted from the "parameter update mirror structure" and denoted as... The formula obtained from formula ① Substitute the deviation term into formula ② In the middle, the calculation is obtained The output of the first part of this section is the procedure summary package and... The "Process Summary Package" is called from the input locations of stage time, cleaning agent dosage, and UV irradiation duration, which are then used to perform pollution risk score updates, reset strategy orchestration, and parameter difference analysis. It is directly read by the risk update subprocess.

[0084] Following the aforementioned intermediate quantities, the main control unit, in the process summary package, the status of the sealed container, and... Based on this, the pollution risk score update and reset strategy arrangement are integrated and judged. For the linkage assessment of resource usage and stage timing, the main control unit adopts a combination of weighted fusion and quadratic cost, and introduces scheduling window constraints to quickly determine the feasibility of strategy implementation. First, the main control unit constructs a session-level score based on the differences between the risk segment parameters of the previous nursing session and the process summary package of the current session, and then corrects it according to the weight of the items requiring attention, forming a write-backable risk update value. Formula ③

[0085]

[0086] in:

[0087] Updated value for session-level contamination risk;

[0088] This is the weighted column vector of the stage resource vector;

[0089] A weighted scalar for completion penalty;

[0090] The container state affects the weight scalar;

[0091] The container status indicator is a combined metric derived from the estimated available container capacity in the sealed container status and the container replacement reminder event marker.

[0092] The mapping from data source to metric to variable is as follows:

[0093] Extract the "estimated available container balance and container replacement prompt event flag" from the "sealed container status" and map them to... The "stage resource weight and penalty weight" are extracted from the "parameter update mirror structure" and denoted as... and The "Container Impact Weight" is extracted from the "Items to Watch Marking Strategy Table" and recorded as... Formula ③ directly uses the formula from formula ①. With formula ② ,get This is then used as a core field in the risk score update record for subsequent calculations. Formula ④

[0094]

[0095] in:

[0096] The reset strategy vector includes the cover plate positioning angle sequence, the support and fall step distance sequence, the spray metering valve reset opening degree, the negative pressure pump shutdown delay, the inlet gate locking sequence, and the ultraviolet irradiation shutdown sequence.

[0097] The feasible domain for resources and time windows is derived from the time window constraints and conflict constraints formed by the execution sequence instruction set and resource occupancy registration.

[0098] Find the vector operator that minimizes the objective function;

[0099] This is the mapping matrix from policy to action domain;

[0100] The target motion trajectory vector originates from the reset target segment in the parameter update mirror structure;

[0101] For risk coupling weights.

[0102] The mapping from data source to metric to variable is as follows: it is composed of "time window constraints and conflict constraints" extracted from "execution sequence instruction set and resource usage registration". It is composed of "reset target fragment and mapping coefficients" extracted from "parameter update mirror structure". and The "risk coupling weight" extracted from the "weight configuration" is denoted as... Formula ④ employs a combined solution approach of quadratic cost and time window constraints (within the scope of operations research and scheduling), within the feasible region. The content is given inside. The output of the second part of this section is the risk score update record and reset strategy vector, which are respectively called by the input positions of the nursing log entry generation and parameter update mirror output operation and the input positions of the cover plate and support reset control. It was issued directly as an execution item.

[0103] Following the aforementioned indicators, the main control unit obtains the risk score update record. With reset strategy vector Then, the two are mapped to executable control variables and revertible configuration variables, and cross-session data accumulation is completed. To construct hardware-readable control sequences and traceable log structures, the main control unit first... Action domain discretization and topological sorting enable actions to form a continuous execution flow on the hardware side under sequential constraints such as shutdown, fallback, halt, latching, and off-lighting. Subsequently, risk and policy elements are extracted as mirror parameter slices to establish the scale, target, and upper limit configuration for the next session. (Formula ⑤)

[0104]

[0105] in:

[0106] Update the vectorized form of the mirror structure for the parameters;

[0107] This is the mapping matrix from the policy to the mirror parameters;

[0108] The compensation matrix for phase resources to mirror parameters;

[0109] The reset strategy vector obtained from formula ④;

[0110] The stage resource vector is obtained from formula ①.

[0111] The mapping from data source to metric to variable is constructed by extracting the mapping coefficients from the "Action to Parameter Mapping Table". It is composed of "compensation coefficients" extracted from the "scale compensation table". The result obtained from formula ④ The formula obtained from formula ① Co-generation ,in It includes the scale configuration, stage goals, and upper limit configuration for the next nursing session. Formula ⑥

[0112]

[0113] in:

[0114] The serialization result of nursing log entries;

[0115] This is a concatenation function;

[0116] This is a serialization function that transcribes control variables and indicators into structured text fragments.

[0117] , , These are the reset strategy vector, risk update value, and stage resource vector, respectively.

[0118] The mapping from data source to metric to variable is as follows:

[0119] The serialized input consists of "Execution Log Records," "Risk Records," and "Stage Resource Vectors." The output of Formula ⑥ is... This refers to the nursing log entries. At this point, the third part of this section outputs a mirror structure of nursing log entries and parameter updates: where the nursing log entries... It is stored externally as a nursing record; parameter update mirror structure. When the next nursing session starts, it is directly invoked by the field dictionary mapping and unit normalization configuration of S100 and the input positions of session number generation, stage state initialization and rule table matching of S200. In summary, this step quantifies and unifies execution results, timing, and resources within a two-level cost and constraint framework, outputs an executable reset strategy and re-feedbackable mirror parameters, and generates a traceable log structure within the same session, forming a closed-loop link for the next session.

Claims

1. A bed pan automatic control method based on multi-sensor fusion, characterized in that, The method comprises: acquiring proximity / posture sensing data, force / support sensing data, and excretion completion sensing data, performing time reference alignment, channel number normalization, and noise segment labeling processing, and extracting proximity trend features, in-place stability indicators, and humidity residual curves, performing field dictionary mapping, dimension normalization, segmented slicing, phase threshold binding, duration statistics, and abnormal segment rejection processing, and generating a phase criterion set; performing session number generation, phase state initialization, and care session binding processing to obtain a care session ID and a phase state, extracting proximity-in-place state, in-place support state, and end determination state, and performing pollution factor quantization, time slice aggregation, and rule table matching processing to generate a pollution risk score initial value, and performing instruction slot arrangement, sequence constraint loading, and resource occupation registration processing to generate an execution sequence instruction set; performing electric cover opening and closing instruction issuing, support positioning instruction issuing, and in-place monitoring subscription, performing spray flushing instruction issuing, vacuum suction instruction issuing, and sealed container inlet isolation control processing, performing ultraviolet irradiation instruction issuing, secondary process trigger determination, and container capacity evaluation processing, and generating disinfection execution records and sealed container states; performing phase timestamp aggregation, instruction completion degree aggregation, and resource consumption registration, extracting phase time, cleaning agent dosage, and ultraviolet irradiation time, and performing pollution risk score updating, reset strategy arrangement, and parameter difference analysis processing, generating reset instructions for the electric cover mechanism and the support positioning mechanism, and risk score update records associated with the care session ID, performing cover and support reset control, care log entry generation, and parameter update mirror output operation, and constructing a care log entry and parameter update mirror structure.

2. The method of claim 1, wherein, The process of the pollution risk score further comprises: The pollution risk score updating is based on the phase time, cleaning agent dosage, and ultraviolet irradiation time fields, and is compared and updated with the previous pollution risk score initial value, and is segmented and adjusted in combination with the degradation execution marker, the item of attention marker, and the container replacement prompt event marker, to generate a pollution risk score update record associated with the care session ID.

3. The method of claim 1, wherein, The reset strategy arrangement comprises: The reset strategy arrangement generates a reset instruction for the electric cover mechanism, the support positioning mechanism, the spray assembly, the vacuum suction assembly, and the sealed container inlet isolation structure according to the item of attention marker and the container replacement prompt event marker, and the reset instruction indicates the replacement, positioning, or resource replenishment state of the reset.

4. The method of claim 1, wherein, The parameter difference analysis processing comprises: The parameter difference analysis processing compares the phase time, cleaning agent dosage, ultraviolet irradiation time, and degradation execution event count with the corresponding fields of the previous care session, identifies the deviation amplitude, and generates a difference segment, and writes the difference segment into the pollution risk score update record.

5. The method of claim 1, wherein, The execution phase timestamp aggregation comprises: The phase timestamp aggregation comprises reading the spray, suction, inlet isolation, and ultraviolet irradiation timestamps and establishing a unified time line index to form a phase time table, which is used for subsequent phase time reconstruction and care log entry generation.

6. The method of claim 1, wherein, The instruction completion degree aggregation comprises: The instruction completion degree aggregates the completion flag, the degradation flag or the exception event buffer flag of each instruction slot in the execution sequence instruction set in the main control unit, forms a completion degree mapping entry and records the items to be concerned.

7. The method of claim 1, wherein, The resource consumption registration includes: The resource consumption registration accumulatively registers the cleaning agent dose, the negative pressure pump working time, the ultraviolet irradiation time and the support driving occupation time, and is associated with the container available amount estimation value and the replacement prompt event mark in the sealed container state.

8. The method of claim 1, wherein, The process of updating the risk score record associated with the nursing session ID also includes: The nursing log entry is generated by the combination of the nursing session ID, the stage schedule, the instruction completion degree mapping entry, the resource consumption registration result, the pollution risk score update record, the container replacement prompt event mark and the cover and support reset state.

9. The method of claim 1, wherein, The process of constructing the nursing log entry and the parameter update mirror structure includes: The parameter update mirror structure encapsulates the risk score update record, the difference segment and the container replacement prompt event mark, and contains the field dictionary update information of the next nursing session, the spray metering valve throttling configuration, the ultraviolet irradiation period configuration and the container capacity determination threshold.

10. The method of claim 1, wherein, The process of constructing the nursing log entry and the parameter update mirror structure also includes: The parameter update mirror structure parses the baseline and the threshold baseline at the start of the next nursing session, is used for proximity / posture sensing data analysis, nursing session binding and pollution risk dynamic adjustment, and realizes cross-session adaptive control closed loop.

Images