Anesthetic drug delivery robot service data processing method and system

Abstract

The present application relates to the technical field of data processing, in particular to a narcotic drug delivery robot business data processing method and system. The method comprises: obtaining a target discrimination index value corresponding to a stagnation period, adjusting the DTW distance between the cumulative path length sequence corresponding to the stagnation period in each actual delivery sub-period and the cumulative path length sequence corresponding to the matching planning delivery path segment of the corresponding actual delivery sub-period according to the target discrimination index value, and obtaining the difference evaluation result between the planning delivery path and the actual delivery path of the target narcotic drug delivery task. Moreover, the present application adjusts the DTW distance between the actual delivery path and the planning delivery path according to the target discrimination index value, which can improve the credibility of the similarity or difference evaluation of the actual delivery path and the planning delivery path of the narcotic drug delivery robot.

Description

Technical Field

[0001] This invention relates to the field of data processing, specifically to a method and system for processing operational data of a robot for transporting anesthetic drugs. Background Technology

[0002] To improve the safety, efficiency, and regulatory oversight of anesthetic drugs within hospitals, anesthetic drug delivery robots have been introduced into the field of anesthetic drug transportation. These robots are intelligent mobile devices deployed within hospitals to conduct closed-loop, traceable, and controlled drug transportation between operating rooms, anesthesiology departments, pharmacy departments, and inpatient wards.

[0003] Currently, to ensure delivery safety and compliance, such as identifying violations or risks like route deviations, abnormal detours, and abnormal stops during actual delivery, it is usually necessary to perform path trajectory verification or analyze and evaluate the similarity or difference between the actual and planned delivery paths of anesthetic drug delivery robots. Path trajectory verification or the evaluation of the similarity or difference between the actual and planned delivery paths of anesthetic drug delivery robots is part of the business data processing for anesthetic drug delivery robots. That is, the business data processing for anesthetic drug delivery robots includes the evaluation of the similarity or difference between the actual and planned delivery paths. Existing methods typically rely on dynamic time warping (DTW) to evaluate the similarity or difference between the actual and planned delivery paths of anesthetic drug delivery robots. However, traditional DTW aligns sequences through elastic scaling of the time axis, and there is a clear uniqueness between the actual and planned delivery paths of anesthetic drug delivery robots. In a one-to-one correspondence scenario, the free alignment of DTW (Dynamic Time Warping) can introduce invalid paths, leading to low reliability of the evaluation results. For example, the robot may exhibit normal pausing behavior, but when the robot is pausing normally, DTW will generate multiple duplicate points at the same location. This repeated matching of points during the normal pausing phase causes a sharp increase in the cumulative distance, resulting in biased or excessive similarity or difference assessment results. Consequently, when identifying violations or risks such as route deviations, abnormal detours, and abnormal stops based on the obtained similarity or difference assessment results, identification errors may occur, such as identifying normal behavior as abnormal behavior or low-risk abnormal behavior as high-risk abnormal behavior. Therefore, when using traditional dynamic time warping to assess the similarity or difference between the actual and planned transport paths of anesthetic drug delivery robots, the obtained similarity or difference assessment results will have low reliability. Thus, improving the reliability of the assessment becomes an urgent issue. Summary of the Invention

[0004] To address the aforementioned problems, this invention provides a method and system for processing operational data of a narcotic drug delivery robot, the specific technical solution of which is as follows:

[0005] In a first aspect, embodiments of the present invention provide a method for processing operational data of an anesthetic drug delivery robot, comprising the following steps:

[0006] The matching planned transportation path segments for each actual transportation sub-time period in the actual transportation time period corresponding to the target anesthetic drug transportation task are obtained, as well as the actual location points corresponding to each actual transportation time in each actual transportation sub-time period and the cumulative path length of each actual operation time, the planned location points on the matching planned transportation path segments, the cumulative path length of each planned location point, and the fixed stop planned location points corresponding to the target anesthetic drug transportation task.

[0007] The stagnation time period in each actual transportation sub-time period is obtained based on the cumulative path length of the actual running time. The target discrimination index value corresponding to the stagnation time period is obtained based on the robot log of each actual running time in the stagnation time period, the distance between the actual position point corresponding to the actual running time in the stagnation time period and the fixed stop planning position point, the difference in cumulative path length between adjacent actual running times in the stagnation time period, and the angle of the vector formed by the actual position points corresponding to adjacent actual running times.

[0008] Based on the target discrimination index value, the DTW distance between the cumulative path length sequence corresponding to the stagnation time period in each actual transportation sub-time period and the cumulative path length sequence corresponding to the matching planned transportation path segment of the corresponding actual transportation sub-time period is adjusted to obtain the difference evaluation result between the planned transportation path and the actual transportation path of the target anesthetic drug transportation task.

[0009] Secondly, embodiments of the present invention provide a business data processing system for an anesthetic drug delivery robot, including a memory and a processor, wherein the processor executes a computer program stored in the memory to implement the aforementioned business data processing method for an anesthetic drug delivery robot.

[0010] Beneficial Effects: This invention first obtains the stagnation time periods in each actual transport sub-time period based on the cumulative path length of the actual running time. Then, based on the robot logs for each actual running time within the stagnation time period, the distance between the actual location point corresponding to the actual running time and the fixed planned stopping location point within the stagnation time period, the difference in cumulative path length between adjacent actual running times within the stagnation time period, and the angle of the vector formed by the actual location points corresponding to adjacent actual running times, a target discrimination index value corresponding to the stagnation time period is obtained. Next, based on the target discrimination index value, the DTW distance between the cumulative path length sequence corresponding to the stagnation time period in each actual transport sub-time period and the cumulative path length sequence corresponding to the matched planned transport path segment of the corresponding actual transport sub-time period is adjusted to obtain the difference assessment result between the planned transport path and the actual transport path for the target anesthetic drug transport task. Furthermore, this invention adjusts the DTW distance between the actual transport path and the planned transport path based on the target discrimination index value, which can improve the reliability of the similarity or difference assessment between the actual transport path and the planned transport path of the anesthetic drug transport robot, thereby improving the accuracy of identifying and judging violations or risky behaviors such as route deviation, abnormal detours, and abnormal stops during actual transport. Attached Figure Description

[0011] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a flowchart of a data processing method for a robot transporting anesthetic drugs according to the present invention. Detailed Implementation

[0013] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention are within the protection scope of the embodiments of the present invention.

[0014] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art.

[0015] This embodiment provides a method for processing operational data from a robot transporting anesthetic drugs, detailed below:

[0016] like Figure 1As shown, the data processing method for the anesthetic drug delivery robot includes the following steps:

[0017] Step S001: Obtain the matched planned transportation path segment for each actual transportation sub-time period in the actual transportation time period corresponding to the target anesthetic drug transportation task, the actual location point corresponding to each actual transportation time in each actual transportation sub-time period and the cumulative path length of each actual operation time, each planned location point on the matched planned transportation path segment, the cumulative path length of each planned location point and the fixed stop planned location point corresponding to the target anesthetic drug transportation task.

[0018] Traditional dynamic time warping (DTW) aligns sequences by elastically scaling the time axis, minimizing the global cost through path search to achieve alignment of similar shapes. However, the actual operating path of an anesthetic drug delivery robot has a clear and unique correspondence with the planned delivery path. In this case, the free alignment of DTW introduces invalid paths, leading to low reliability of the evaluation results. For example, the anesthetic drug delivery robot may exhibit normal pausing behavior during operation. However, when the robot pauses normally, DTW generates multiple duplicate points at the same location, meaning that points during the normal pausing phase are repeatedly matched, causing a sharp increase in the cumulative distance and consequently affecting the obtained results. If the similarity or difference assessment results deviate or are too large, subsequent identification of violations or risks such as route deviations, abnormal detours, and abnormal stops based on the obtained similarity or difference assessment results may result in identification errors, such as identifying normal behavior as abnormal behavior or low-risk abnormal behavior as high-risk abnormal behavior. In order to improve the credibility of the assessment, this embodiment will subsequently match and align the planned transportation route with the actual transportation route based on the necessary business operations of anesthetic drug transportation to avoid introducing invalid routes during alignment. Then, the stalls in each segment of the actual transportation route will be distinguished and the DTW distance will be adjusted to improve the credibility of the assessment.

[0019] In addition, anesthetic drug delivery robots typically have the following characteristics: During maintenance, no one without the appropriate area access is allowed to operate the robot, and its structural design ensures unauthorized access. The robots also strictly adhere to anesthetic drug management procedures, displaying real-time inventory quantities and expiration dates of each drug, following the principle of using drugs nearing their expiration date first, providing automatic warnings when drugs are nearing their expiration date, marking expired drugs as disabled, and locking their release access. The anesthetic drug delivery robot's drug retrieval process employs dual authentication: two people verify their identities and permissions. One person is the borrower, and the other must be registered in the system as a doctor, pharmacist, or nurse. The robot unlocks only when both authentication signals arrive simultaneously, ensuring proper storage of anesthetic drugs.

[0020] Based on the above description and for ease of understanding, this embodiment will subsequently describe the process of evaluating the similarity or difference between the planned and actual transport routes of any anesthetic drug transport task, and the anesthetic drug transport task will be referred to as the target anesthetic drug transport task.

[0021] Since the anesthetic drug delivery robot performs path planning before performing the delivery task, this embodiment first obtains the planned path when performing the target anesthetic drug delivery task and records it as the planned delivery path of the target anesthetic drug delivery task. After obtaining the planned delivery path of the target anesthetic drug delivery task, the planned position points on the planned delivery path are obtained by sampling. The distance between adjacent planned position points on the planned delivery path can be set according to the total length of the planned delivery path, the robot's travel speed, and other actual conditions. For example, the distance between adjacent planned position points can be set to the length traveled per second by the robot when performing the target anesthetic drug delivery task.

[0022] Since the route planning includes the setting of data such as fixed stop points and business operations, in this embodiment, the fixed stop points set on the planned transportation route of the target anesthetic drug transportation task will be recorded as the fixed stop planned position points corresponding to the target anesthetic drug transportation task. The coordinates of all position points in this embodiment are determined by the same two-dimensional coordinate system. The horizontal axis and vertical axis of the two-dimensional coordinate system can be determined according to the actual situation, such as choosing the horizontal direction to the right as the positive direction of the horizontal axis and the vertical direction to the up as the positive direction of the vertical axis. Then, the business operations specified or required when planning the target anesthetic drug delivery task are obtained, and all of them are recorded as necessary business operations for the target anesthetic drug delivery task. If the business operations that the robot must perform when planning the target anesthetic drug delivery task include opening the cabin and triggering the access control, then in this embodiment, the necessary business operations for the target anesthetic drug delivery task are opening the cabin and triggering the access control. That is, there will be necessary triggering business operations when planning the target anesthetic drug delivery task, or certain location points will be required or specified to trigger fixed business operations when planning the delivery path. For example, it is specified that the cabin opening operation will be triggered at a certain location point for drug retrieval. Opening the cabin refers to the operation of opening the drug storage compartment of the anesthetic drug delivery robot. The access control triggering when the anesthetic drug delivery robot performs the task refers to the process in which the robot automatically controls the electromagnetic lock to be de-energized and released to open the door through the linkage of the sensor or communication system with the access control device. Necessary business operations are the key to matching and aligning the planned delivery path with the actual delivery path and avoiding the introduction of invalid paths by free alignment. Fixed stopping planning location points are key parameters for distinguishing the stagnation time period and adjusting DTW.

[0023] After obtaining the planned transport path, planned location points on the planned transport path, fixed stop planned location points corresponding to the target anesthetic drug transport task, and necessary business operations for the target anesthetic drug transport task, this embodiment obtains the actual transport time period when executing the target anesthetic drug transport task, and records it as the actual transport time period corresponding to the target anesthetic drug transport task. In this embodiment, the actual transport time period refers to the time period from when the anesthetic drug transport robot starts executing the target anesthetic drug transport task to when it completes the target anesthetic drug transport task. After obtaining the actual transport time period corresponding to the target anesthetic drug transport task, the actual location points corresponding to each actual running time in the actual transport time period are obtained based on the data recorded by the anesthetic drug transport robot. The coordinate value of the actual location point corresponding to any actual running time refers to the coordinate value of the position of the anesthetic drug transport robot in the two-dimensional coordinate system at that actual running time.

[0024] After obtaining the actual location points corresponding to each actual running time and the planned location points on the planned transportation route within the actual transportation time period corresponding to the target anesthetic drug delivery task, the cumulative path length for each actual running time and the cumulative path length for each planned location point are obtained. The cumulative path length is a key parameter for subsequent differentiation of stall time periods and DTW adjustment. The specific process for obtaining the cumulative path length is as follows:

[0025] First, the cumulative path length of each planned location point on the planned transportation route is obtained. The cumulative path length of any planned location point refers to the total path length from the starting point on the planned transportation route to that planned location point. Then, on the planned transportation route, the planned location point closest to the actual location point corresponding to each actual transportation time is obtained and recorded as the nearest planned location point corresponding to the corresponding actual operation time. For example, if the Euclidean distance between planned location point a0 and the actual location point corresponding to actual transportation time t0 is the smallest among all planned location points on the planned transportation route, then planned location point a0 is the nearest planned location point corresponding to actual transportation time t0. Then, the cumulative path length of the nearest planned location point corresponding to each actual operation time is used as the cumulative path length of the corresponding actual operation time. This process transforms two-dimensional coordinates into one-dimensional parameters, which is convenient for subsequent DTW calculations. That is, this process is to unify the reference coordinates of the planned transportation route and the actual transportation route, using the cumulative path length mapped to the planned transportation route as a standardization reference.

[0026] Next, this embodiment divides and matches the planned transportation route and the actual transportation time period based on the known necessary business operations of the target anesthetic drug transportation task, to obtain the matched planned transportation route segments for each actual transportation sub-time period in the actual transportation time period corresponding to the target anesthetic drug transportation task. The specific process of obtaining the matched planned transportation route segments for each actual transportation sub-time period in the actual transportation time period corresponding to the target anesthetic drug transportation task is as follows:

[0027] First, the locations of the necessary operational points that trigger the target anesthetic drug delivery task along the planned delivery route are obtained and recorded as necessary locations on the planned delivery route. During the actual delivery time, the times when the necessary operational points that trigger the target anesthetic drug delivery task are obtained and recorded as necessary moments. For example, according to the logs recorded by the anesthetic drug delivery robot, the robot triggers a necessary operational point for the target anesthetic drug delivery task at a certain moment during the execution of the task. If this moment is time t1 within the actual delivery time, then time t1 is a necessary moment. Triggering a necessary operational point at a certain location or time means that the robot automatically executes operational operations related to anesthetic drug transportation when it reaches the preset location or time node. For example, if the robot triggers the hatch opening operation at a certain actual delivery moment, that actual delivery moment is a necessary moment. Then, the actual delivery time is segmented using the necessary moments within the actual delivery time to obtain each actual delivery sub-time period. The time period formed by adjacent necessary moments within a segment is an actual transportation sub-segment. Alternatively, the first actual transportation sub-segment is the time period formed by the first necessary moment and the second necessary moment within the actual transportation sub-segment. Furthermore, if the start and end times of an actual transportation sub-segment are not necessary moments, then the start time and the necessary moment closest to the start time will also form an actual transportation sub-segment, and the end time and the necessary moment closest to the end time will also form an actual transportation sub-segment. Then, based on the necessary moments in each actual transportation sub-segment and the necessary location points on the planned transportation path, the matching planned transportation path segments for each actual transportation sub-segment are obtained. The specific process for obtaining the matching planned transportation path segments for each actual transportation sub-segment based on the necessary moments in each actual transportation sub-segment and the necessary location points on the planned transportation path is as follows:

[0028] For any actual transportation sub-time period A, which consists of two adjacent necessary moments: The two necessary moments in actual transportation sub-time period A are designated as the first moment and the second moment, respectively. The necessary location points that are identical to the necessary business operation triggered by the first moment are identified and recorded as the necessary location points corresponding to the first moment. That is, the necessary business operation triggered by the necessary location point corresponding to the first moment is the same as the necessary business operation triggered by the first moment. Among all the necessary location points corresponding to the first moment, the necessary location point closest to the actual location point corresponding to the first moment is identified and recorded as the matching location point of the first moment. That is, among all the necessary location points corresponding to the first moment, the Euclidean distance between the matching location point of the first moment and the actual location point corresponding to the first moment is the smallest. The necessary location points that are identical to the necessary business operation triggered by the second moment are identified and recorded as the necessary location points corresponding to the second moment. Among all the necessary location points corresponding to the second moment, the necessary location point closest to the actual location point corresponding to the second moment is identified and recorded as the matching location point of the second moment. The method for obtaining the matching location point is the same as that for obtaining the matching location point at the first moment. On the planned transportation path, the planned transportation path segment from the matching location point at the first moment to the matching location point at the second moment is obtained and used as the matching planned transportation path segment for the actual transportation sub-time period A. In addition, since the robot will strictly follow the business operations specified before performing the anesthetic drug transportation task, if the robot is specified or planned to open the hatch four times when performing the task, it will usually not open the hatch three or five times in actual execution. Therefore, if an actual transportation sub-time period A contains the start time of the actual transportation time period, then the matching planned transportation path segment for that actual transportation sub-time period is the planned transportation path segment from the start point on the planned transportation path to the first necessary location point on the planned transportation path. If an actual transportation sub-time period contains the start time of the end time of the actual transportation time period, then the matching planned transportation path segment for that actual transportation sub-time period is the planned transportation path segment from the last necessary location point on the planned transportation path to the end point on the planned transportation path.For example, if the prescribed or planned sequence of necessary operational operations for delivering a target anesthetic drug is opening the hatch, access control triggering, and so on, and the necessary locations on the planned delivery route are not the start and end points, and the start and end times in the actual delivery period are not necessary times, but the first necessary location on the planned delivery route triggers the first necessary operational operation (hatch opening), the second necessary location triggers the second necessary operational operation (access control triggering), and the third necessary location triggers the third necessary operational operation (hatch opening). The necessary business operation involves opening the cargo hold. The first necessary moment in the actual transportation time period triggers the first necessary business operation opening in the necessary business operation time sequence; the second necessary moment triggers the second necessary business operation access control trigger; and the third necessary moment triggers the third necessary business operation opening in the necessary business operation time sequence. Therefore, the first actual transportation time segment in the actual transportation time period consists of the period from the start time of the actual transportation time period to the first necessary moment. The matched planned transportation path segment for the first actual transportation time segment in the actual transportation time period is the matched planned transportation path segment. The planned transportation path segment from the starting point on the delivery route to the first necessary location point on the matched planned transportation path; the second actual transportation time segment in the actual transportation time period consists of the period from the first necessary moment to the second necessary moment in the actual transportation time period; the matched planned transportation path segment for the second actual transportation time segment in the actual transportation time period is the planned transportation path segment from the first necessary location point on the matched planned transportation path to the second necessary location point on the matched planned transportation path; the third actual transportation time segment in the actual transportation time period consists of the period from the second necessary moment to the actual transportation time. The third necessary moment in the segment constitutes the planned transportation path segment of the third actual transportation time segment in the actual transportation time period, which is the planned transportation path segment from the second necessary location point on the planned transportation path to the third necessary location point on the planned transportation path. The fourth actual transportation time segment in the actual transportation time period consists of the third necessary moment in the actual transportation time period to the end moment of the actual transportation time period. The planned transportation path segment of the fourth actual transportation time segment in the actual transportation time period is the planned transportation path segment from the third necessary location point on the planned transportation path to the end point on the planned transportation path.Furthermore, the robot delivery plan itself includes notes and plans for necessary business operations. That is, the planned delivery route has clear locations that trigger necessary business operations. At the same time, the time when necessary business operations are triggered can also be located through robot logs during actual delivery. Therefore, this embodiment aligns the actual delivery time period with the planned delivery route based on the necessary business operations. That is, it obtains the matching planned delivery route segment of the actual delivery sub-time period. It does not rely on the free alignment of DTW, which can fundamentally avoid the introduction of invalid paths. In other words, by matching here and then distinguishing the type of stagnation time and adjusting DTW, it can minimize the problem of large DTW distance deviation caused by repeated matching of normal stagnation intervals, or reduce the problem of low credibility of similarity or difference evaluation results caused by normal stagnation intervals.

[0029] Therefore, this embodiment can obtain the matched planned transportation path segments for each actual transportation sub-time period in the actual transportation time period corresponding to the target anesthetic drug transportation task, the actual location points corresponding to each actual transportation time in each actual transportation sub-time period, the cumulative path length of each actual operation time, each planned location point on the matched planned transportation path segment, the cumulative path length of each planned location point, the fixed stop planned location points corresponding to the target anesthetic drug transportation task, and the corresponding necessary business operations, etc.

[0030] Step S002: Obtain the stagnation time period in each actual transportation sub-time period based on the cumulative path length of the actual running time. Based on the robot logs of each actual running time in the stagnation time period, the distance between the actual location point corresponding to the actual running time in the stagnation time period and the fixed stop planning location point, the difference in cumulative path length between adjacent actual running times in the stagnation time period, and the angle of the vector formed by the actual location points corresponding to adjacent actual running times, obtain the target discrimination index value corresponding to the stagnation time period.

[0031] After completing the matching and alignment, that is, after obtaining the matched planned transportation path segment for the actual transportation sub-time period, this embodiment obtains the stagnation period for each actual transportation sub-time period. The purpose of obtaining the stagnation period is to reduce the cumulative distance expansion caused by repeated matching of normal stagnation intervals. This embodiment then needs to obtain the stagnation period in each actual transportation sub-time period based on the cumulative path length of each actual transportation moment in each actual transportation sub-time period. The specific process of obtaining the stagnation period in each actual transportation sub-time period based on the cumulative path length of each actual transportation moment in each actual transportation sub-time period is as follows: For any actual transportation sub-time period A:

[0032] First, the feature points corresponding to each actual running time in the actual transportation sub-time period A are obtained. The x-coordinate of the feature point corresponding to any actual running time is the corresponding actual running time, and the y-coordinate is the cumulative path length of the corresponding actual running time. Then, based on the absolute value of the slope between the feature points corresponding to adjacent actual running times in the actual transportation sub-time period A, the slope feature value sequence corresponding to the actual transportation sub-time period A is obtained. The v-th slope feature value in the slope feature value sequence corresponding to the actual transportation sub-time period A is the absolute value of the slope between the feature point corresponding to the v-th actual running time and the feature point corresponding to the (v+1)-th actual running time in the actual transportation sub-time period A. Moreover, the actual running time corresponding to the v-th slope feature value in the slope feature value sequence corresponding to the actual transportation sub-time period A is the actual running time between the v-th and (v+1)-th actual running times in the actual transportation sub-time period A. The process of obtaining the slope between the two points is a known technique. Then, the stagnation time periods in the actual transportation sub-time period A are obtained based on the slope feature value sequence corresponding to the actual transportation sub-time period A. The specific process of obtaining the stagnation time periods in the actual transportation sub-time period A based on the slope feature value sequence corresponding to the actual transportation sub-time period A is as follows:

[0033] First, obtain the stagnation confidence representation value corresponding to each slope feature value in the slope feature value sequence corresponding to the actual transportation sub-time period A. Subtracting 0 from any slope feature value and performing a negative correlation mapping yields the stagnation confidence representation value corresponding to that slope feature value. Here, a negative exponential function with a base of constant e is used for the negative correlation mapping, i.e., the stagnation confidence representation value corresponding to the i-th slope feature value in the slope feature value sequence is... , Let be the i-th slope eigenvalue in the slope eigenvalue sequence, and exp() be an exponential function with base e. express The smaller the difference compared to when the slope is completely stationary (i.e., when the slope is 0), the better. The more likely the corresponding time is to begin stagnation; then, identify the slope feature value in the slope feature value sequence where the first stagnation confidence value is greater than the preset confidence threshold, and start the stagnation confidence value judgment process from the slope feature value in the identified slope feature value sequence where the first stagnation confidence value is greater than the preset confidence threshold, to obtain all stagnation time periods in the actual transportation sub-time period A, and the specific explanation of starting the stagnation confidence value judgment process from the slope feature value sequence where the first stagnation confidence value is greater than the preset confidence threshold, to obtain all stagnation time periods in the actual transportation sub-time period A is as follows:

[0034] If the i-th slope feature value in the slope feature value sequence is the first slope feature value in the sequence whose stagnation confidence value is greater than a preset confidence threshold, then the i-th slope feature value is recorded as the initial slope feature value. It is then determined whether the stagnation confidence value corresponding to the (i+1)-th slope feature value in the slope feature value sequence is greater than the preset confidence threshold. If it is, the process continues to determine whether the stagnation confidence value corresponding to the (i+2)-th slope feature value in the slope feature value sequence is greater than the preset confidence threshold. If it is, the process continues to determine whether the stagnation confidence value corresponding to the (i+3)-th slope feature value in the slope feature value sequence is greater than the preset confidence threshold. If it is not greater, the i-th slope feature value in the slope feature value sequence is... The time interval from the first slope feature value to the (i+2)th slope feature value, corresponding to the actual running time, is denoted as the stagnation period in the actual transportation sub-time period A. Following the (i+2)th slope feature value, we continue to acquire slope feature values ​​whose initial stagnation confidence representation value is greater than a preset confidence threshold. These slope feature values ​​are then used as new initial slope feature values, and the stagnation confidence representation value judgment process continues until the entire slope feature value sequence is traversed. This process yields all stagnation periods in the actual transportation sub-time period. In other words, the time intervals corresponding to consecutive slope feature values ​​in the slope feature value sequence where the stagnation confidence representation value is greater than the preset confidence threshold are the stagnation periods in the actual transportation sub-time period. In practical applications, implementers need to set the preset confidence threshold based on the actual situation, such as the range of stagnation confidence representation values. For example, since the range of stagnation confidence representation values ​​is 0 to 1, the preset confidence threshold can be set to 0.5.

[0035] Example of obtaining all stagnation periods in actual transportation sub-time period A: If the slope feature value that first appears in the slope feature value sequence corresponding to the actual transportation sub-time period is greater than the preset confidence threshold, then the slope feature value is the second slope feature value in the slope feature value sequence. If the stagnation confidence values ​​corresponding to the second to fourth slope feature values ​​are all greater than the stagnation confidence value, the stagnation confidence value corresponding to the fifth slope feature value is not greater than the stagnation confidence value, and the stagnation confidence values ​​corresponding to the sixth to eighth slope feature values ​​are all greater than the stagnation confidence value, then the time period formed by the times corresponding to the second to fourth slope feature values ​​is a stagnation period in actual transportation sub-time period A, and the time period formed by the times corresponding to the sixth to eighth slope feature values ​​is a stagnation period in actual transportation sub-time period A.

[0036] Therefore, this embodiment can obtain the stagnation time periods in the actual transportation sub-time periods. However, there will be abnormal stagnation time periods and normal stagnation time periods. Abnormal stagnation time periods refer to time periods caused by abnormal factors, which are business events not included in the planning, such as stagnation caused by visual sensor or drive system malfunctions, or forced obstruction. Normal stagnation time periods are caused by normal factors or normal business events. Normal factors and normal business events refer to business events or behaviors planned or established to achieve the target anesthetic drug transportation task. For example, if the business events planned by the implementer to achieve the target anesthetic drug transportation task include drug retrieval, then drug retrieval is a normal business event, and drug retrieval will cause stagnation. However, the stagnation time period caused by drug retrieval is a normal stagnation time period. The phenomenon of DTW calculation results being too large due to normal stagnation will lead to lower reliability of the obtained similarity or difference assessment results. Furthermore, if the stagnation caused by normal stagnation is not resolved or eliminated... The overestimation of DTW calculation results can lead to errors in identifying violations or risks such as route deviations, abnormal detours, and abnormal stops based on the obtained similarity or difference assessment results. To improve the reliability of the similarity or difference assessment results, this embodiment will distinguish between stoppage periods in the actual transport sub-time period based on the difference between normal and abnormal stoppages. Specifically, this embodiment will first obtain the target discrimination index value corresponding to the stoppage period based on the robot logs at each actual running time within the stoppage period, the distance between the actual location point corresponding to the actual running time within the stoppage period and the fixed stoppage planned location point, the cumulative path length difference between adjacent actual running times within the stoppage period, and the angle of the vector formed by the actual location points corresponding to adjacent actual running times. The target discrimination index value is the basis for classifying normal and abnormal stoppages in the region. The specific process for obtaining the target discrimination index value corresponding to any stoppage period Q is as follows:

[0037] First, the normal business events corresponding to the target anesthetic drug delivery task are obtained. Normal business events corresponding to an anesthetic drug delivery task refer to all standardized normal business events during the execution of the anesthetic drug delivery task. In other words, normal business events corresponding to an anesthetic drug delivery task refer to the business events planned or specified by the implementer to achieve or complete the anesthetic drug delivery task. Furthermore, the normal business events corresponding to any anesthetic drug delivery task are usually set according to actual business needs. Common normal business events include, but are not limited to, opening the cabin, locking the cabin, retrieving the drug, and waiting for the elevator. Then, based on the normal business events corresponding to the target anesthetic drug delivery task, the robot logs at each actual running moment in the stagnation period Q, and the distance between the actual location point corresponding to the actual running moment in the stagnation period Q and the fixed planned stopping location point, the first discriminant index value of the stagnation period Q is obtained. The robot logs at the actual running moment refer to the robot logs generated at the corresponding actual running moment. The robot logs can be obtained from the database storing robot data. The specific process for obtaining the first discriminant index value of the stagnation period Q is as follows:

[0038] Based on the robot logs at each actual running moment within the stagnation period Q and the normal business events corresponding to the target anesthetic drug delivery task, feature label values ​​are obtained for each actual running moment within the stagnation period Q. If the information recorded in the robot log at any actual running moment includes a normal business event corresponding to the target anesthetic drug delivery task, or if the robot log at any actual running moment contains a record related to a normal business event corresponding to the target anesthetic drug delivery task, then 1 is used as the feature label value for that actual running moment. Conversely, if the information recorded in the robot log at any actual running moment does not include a normal business event corresponding to the target anesthetic drug delivery task, or if the robot log at any actual running moment does not contain a record related to a normal business event corresponding to the target anesthetic drug delivery task, then 0 is used as the feature label value for that actual running moment. The cumulative result of the feature label values ​​corresponding to all actual running moments within the stagnation period Q is calculated, and the cumulative result is then normalized. The result is denoted as the first characteristic value of the stagnation period Q, which is normalized using the hyperbolic tangent function. The average coordinates of the actual location points corresponding to all actual running times in the stagnation period Q are calculated, which is the average location point of all actual running times in the stagnation period Q, and denoted as the representative location point of the stagnation period Q. Among all fixed-stop planned location points, the fixed-stop planned location point closest to the representative location point of the stagnation period Q is obtained and denoted as the nearest neighbor planned location point of the stagnation period Q. The Euclidean distance between the representative location point of the stagnation period Q and the nearest neighbor planned location point of the stagnation period Q is calculated, and a negative correlation mapping is performed on the Euclidean distance. The mapping result is denoted as the second characteristic value of the stagnation period Q, which is negatively correlated using a negative exponential function with a base of constant e. The mean of the first characteristic value and the second characteristic value of the stagnation period Q is calculated and denoted as the first discriminant index value of the stagnation period Q. The expression for calculating the first discriminant index value of the stagnation period Q is:

[0039]

[0040] Where P1 is the first discriminant index value of the stagnation time period Q, tanh() is the hyperbolic tangent function, mainly used for normalization, exp() is an exponential function with a base of constant e, and D0 is the Euclidean distance between the representative location point of stagnation time period Q and the nearest neighbor planned location point of stagnation time period Q. T0 represents the feature label value corresponding to the t-th actual running time in the stagnation period Q, and T0 represents the total number of actual running times in the stagnation period Q. The larger the value of D0, the more likely the standstill period Q is caused by normal business events or activities related to the delivery of the target anesthetic drugs; conversely, the smaller the value of D0, the closer the location point during the standstill period is to these known fixed stop locations, and the more likely the standstill period Q is caused by these normal business activities. The larger P1 is and the smaller D0 is, the larger P1 is. Therefore, when P1 is larger, the stagnation period Q is more likely to be caused by normal business behavior or normal business events, and the stagnation period Q is more likely to be a normal stagnation period. Conversely, the smaller P1 is, the less likely the stagnation period Q is to be caused by normal business behavior or normal business events, and the stagnation period Q is more likely to be an abnormal stagnation period.

[0041] Next, obtain the vector angle sequence corresponding to the stagnation time period Q. The number of data points in the vector angle sequence corresponding to the stagnation time period Q is the total number of actual running moments in the stagnation time period Q minus 1. The f-th vector angle in the vector angle sequence corresponding to the stagnation time period Q refers to the angle between the actual position point corresponding to the f-th actual running moment in the stagnation time period Q and the actual position point corresponding to the (f+1)-th actual running moment in the stagnation time period Q. For example, if the coordinates of the actual position point corresponding to the f-th actual running moment are (x1, y1) and the coordinates of the actual position point corresponding to the (f+1)-th actual running moment are (x2, y2), x1 and x2 are the x-coordinates of the position points, and y1 and y2 are the y-coordinates of the position points. Let the vector between the actual position point corresponding to the (f+1)th actual running time be (x2-x1, y2-y1). Then, the result of the calculation of arctan2(x2-x1, y2-y1) is the angle between the vectors formed by the actual position points corresponding to the f-th and (f+1)th actual running times. arctan2() is a function to calculate the arctangent value. Based on the cumulative path length difference between adjacent actual running times in the stagnation period Q and the vector angle sequence corresponding to the stagnation period Q, the second discriminant index value of the stagnation period Q is obtained. The specific process of obtaining the second discriminant index value of the stagnation period Q is as follows:

[0042] The set of absolute differences between the cumulative path lengths of all adjacent actual running times within a stagnation period Q is denoted as the cumulative path length difference sequence corresponding to stagnation period Q. The h-th cumulative path length difference in this sequence is the absolute difference between the cumulative path length of the h-th actual running time and the cumulative path length of the (h+1)-th actual running time within stagnation period Q. Similarly, the set of absolute differences between adjacent vector angles within the vector angle sequence corresponding to stagnation period Q is denoted as the angle difference sequence corresponding to stagnation period Q. The r angle differences are the absolute values ​​of the differences between the r-th vector angle and the (r+1)-th vector angle in the vector angle sequence corresponding to the stagnation period Q. The result of summing all data in the cumulative path length difference sequence and then performing a negative correlation mapping is denoted as the first mapping value of the stagnation period Q. Similarly, the result of summing all data in the angle difference sequence corresponding to the stagnation period Q and then performing a negative correlation mapping is denoted as the second mapping value of the stagnation period Q. Here, a negative exponential function with a base e is used for the negative correlation mapping. The mean of the first and second mapping values ​​of the stagnation period Q is denoted as the second discriminant index value of the stagnation period Q. The specific calculation expression for the second discriminant index value of the stagnation period Q is as follows:

[0043]

[0044] Where P2 is the second discriminant value for the stagnation period Q, and H is the total number of data points in the cumulative path length difference sequence corresponding to the stagnation period Q. Let be the h-th cumulative path length difference in the cumulative path length difference sequence corresponding to the stagnation period Q, and R be the total number of data points in the angle difference sequence corresponding to the stagnation period Q. Let be the r-th angle difference in the sequence of angle differences corresponding to the stagnation period Q. Since normal stagnation is a normal instruction or business behavior in the execution of a task, the robot is almost completely stationary when the drive system does not execute a forward instruction. Abnormal stagnation, however, is often caused by unexpected factors, so the robot often exhibits repetitive motion during abnormal stagnation. Therefore, when... The smaller the value, the more likely the robot is to be completely stationary during the pause period Q, indicating that the pause period Q is more likely to be caused by normal business activities or events, or in other words, the pause period Q is more likely to be a normal pause; while when... A larger value for Q indicates a significant or frequent change in the robot's direction of movement during that time period. This suggests that the robot exhibits repetitive movement characteristics during the stationary period Q, making it less likely that the stationary period Q is caused by normal business activities or events, or more likely, an abnormal stationary period. The smaller the value of Q, the less frequently the robot's movement direction changes during that time period. A smaller value indicates that the robot's stationary time period Q does not exhibit repetitive movement characteristics, making it more likely that the stationary time period Q is caused by normal business operations, or in other words, it is more likely that the stationary time period Q is a normal stationary period. smaller and The smaller the value of P2, the larger the value of P2. Therefore, when P2 is larger, the stagnation period Q is more likely to be a normal stagnation period. Conversely, the smaller the value of P2, the more likely the stagnation period Q is to be an abnormal stagnation period.

[0045] Finally, the average of the first and second discriminant index values ​​for the stagnation period Q is calculated and denoted as the target discriminant index value corresponding to the stagnation period Q. The larger the target discriminant index value corresponding to the stagnation period Q, the more it indicates that there is a clear normal business behavior or event in the stagnation period Q and that the stopping position is closer to the fixed stop planning position point in the planned operation path in terms of spatial location. At the same time, there is almost no displacement or directional jitter, which also indicates that the stagnation period Q is more likely to be a normal stagnation period or a normal stagnation period. Conversely, the smaller the target discriminant index value corresponding to the stagnation period Q, the more likely the stagnation period Q is to be an abnormal stagnation period.

[0046] Therefore, this embodiment can obtain the target discrimination index value corresponding to any stagnation time period through the above process.

[0047] Step S003: Based on the target discrimination index value, adjust the DTW distance between the cumulative path length sequence corresponding to the stagnation time period in each actual transportation sub-time period and the cumulative path length sequence corresponding to the matching planned transportation path segment of the corresponding actual transportation sub-time period, to obtain the difference evaluation result between the planned transportation path and the actual transportation path of the target anesthetic drug transportation task.

[0048] In this embodiment, after obtaining the target discrimination index value corresponding to the stagnation period, the DTW distance between the cumulative path length sequence corresponding to the stagnation period in the actual transportation sub-time period and the cumulative path length sequence corresponding to the planned transportation path segment of the corresponding actual transportation sub-time period is adjusted based on the target discrimination index value. This yields an evaluation result of the difference between the planned transportation path and the actual transportation path for the target anesthetic drug transportation task. The specific process is as follows:

[0049] The system determines whether the target discrimination index value corresponding to the stagnation period is greater than a preset discrimination threshold. If so, the corresponding stagnation period is classified as a normal stagnation period. Normal stagnation periods can lead to an overestimation of the DTW distance, resulting in unreliable evaluation results, thus requiring adjustment. Therefore, stagnation periods exceeding the preset discrimination threshold are recorded as periods requiring adjustment. Otherwise, the corresponding stagnation period is classified as an abnormal stagnation period. The DTW distance calculated for abnormal stagnation periods does not affect the reliability of the final evaluation result, so no adjustment is needed. Furthermore, in practical applications, implementers need to set the preset discrimination threshold based on actual conditions such as the range of target discrimination index values. For example, in this embodiment, the preset discrimination threshold can be set to 0.5.

[0050] Next, based on the target discrimination index value corresponding to the stagnation period, the target distance of the actual transportation sub-time period is obtained. Specifically, for any actual transportation sub-time period, there is a time period that needs to be adjusted:

[0051] First, calculate the actual sequence corresponding to any time segment within the actual transportation sub-time period. The actual sequence corresponding to any time segment is the sequence formed by the cumulative path lengths of all actual transportation moments within that time segment. The cumulative path length of the y-th actual transportation moment within any time segment is the y-th cumulative path length in the actual sequence corresponding to that time segment. The matching sequence follows the same principle. This actual transportation sub-time period usually only has time segments to be adjusted and non-adjusted time segments. Non-adjusted time segments refer to the time segments within the actual transportation sub-time period other than the time segments to be adjusted. The sequence formed by the cumulative path lengths of all planned location points on the matched planned transportation path segment of this actual transportation sub-time period is denoted as the matching sequence of this actual transportation sub-time period. Calculate the DTW distance between the actual sequence corresponding to the time segment to be adjusted and the matching sequence corresponding to this actual transportation sub-time period, and denot it as the initial distance of the corresponding time segment to be adjusted. The calculation process of the DTW distance between any two sequences is well known. Multiply the initial distance of the time segment to be adjusted by the corresponding time segment to be adjusted. The product of the adjustment degree values ​​of the segments is denoted as the target distance of the corresponding time period to be adjusted. The adjustment degree value of any time period to be adjusted is the negative correlation mapping result of the target discrimination index value corresponding to the time period to be adjusted. Here, the negative correlation mapping refers to the constant 1 minus the target discrimination index value. That is, the adjustment degree value of any time period to be adjusted is the result of the constant 1 minus the target discrimination index value corresponding to the time period to be adjusted. The DTW distance between the actual sequence corresponding to the non-adjusted time period in the actual transportation sub-time period and the matching sequence corresponding to the actual transportation sub-time period is calculated and directly denoted as the target distance of the corresponding non-adjusted time period. The sum of the target distances of all non-adjusted time periods and all target distances of the time periods to be adjusted in the actual transportation sub-time period is calculated and denoted as the target distance of the actual transportation sub-time period. That is, if there is one non-adjusted time period and one time period to be adjusted in the actual transportation sub-time period, then the result of the target distance of the non-adjusted time period and the target distance of the time period to be adjusted in the actual transportation sub-time period is the target distance of the actual transportation sub-time period. This embodiment adjusts the DTW distance between the actual sequence corresponding to the time period to be adjusted and the matching sequence corresponding to the actual transportation sub-time period based on the target discrimination index value, which can improve the credibility of subsequent difference assessment results.

[0052] In addition, the slope feature value sequence corresponding to the actual transportation sub-time period may have some values ​​that are all less than the preset confidence threshold, or there may be some values ​​that are greater than the preset confidence threshold but are not continuous. In such cases, the DTW distance between the sequence formed by the cumulative path length of all actual transportation times in the actual transportation sub-time period and the sequence formed by the cumulative path length of all planned location points on the matched planned transportation path segment of the actual transportation sub-time period can be used as the target distance of the actual transportation sub-time period.

[0053] After obtaining the target distance for the actual transportation sub-time period, the sum of the target distances for all actual transportation sub-time periods within the actual transportation period is used as the difference assessment result between the planned transportation path and the actual transportation path for the target anesthetic drug transportation task. That is, the larger the sum of the target distances for all actual transportation sub-time periods within the actual transportation period, the greater the difference or the smaller the similarity between the planned and actual transportation paths for the target anesthetic drug transportation task. This embodiment can subsequently achieve a comprehensive assessment and risk positioning of the robot transportation business process based on the difference assessment result between the planned and actual transportation paths for the target anesthetic drug transportation task, providing high-precision process monitoring, anomaly location, and performance evaluation for robot transportation operations. For example, it can accurately identify violations or risky behaviors such as route deviation, abnormal detours, and abnormal stops during actual transportation, thereby effectively improving scheduling efficiency and business risk controllability. Furthermore, after completing the difference assessment between the planned and actual transportation paths for the target anesthetic drug transportation task, how to identify and determine violations or risky behaviors such as route deviation, abnormal detours, and abnormal stops during actual transportation based on the assessment results is a known technology.

[0054] This embodiment of a data processing system for a narcotic drug delivery robot includes a memory and a processor. The processor executes a computer program stored in the memory to implement the aforementioned data processing method for a narcotic drug delivery robot.

[0055] Thus, this embodiment completes the assessment of the similarity or difference between the actual and planned transport paths of the anesthetic drug transport robot. In this embodiment, the planned and actual transport paths are matched and aligned based on the necessary business operations for anesthetic drug transport to avoid introducing invalid paths during alignment. Then, the stall segments in each actual transport path are distinguished and DTW adjusted, which can improve the credibility of the similarity or difference assessment between the actual and planned transport paths of the anesthetic drug transport robot.

[0056] In summary, this embodiment first obtains the stagnation time periods in each actual transportation sub-time period based on the cumulative path length of the actual running time. Then, based on the robot logs for each actual running time within the stagnation time period, the distance between the actual location point corresponding to the actual running time and the fixed planned stopping location point within the stagnation time period, the difference in cumulative path length between adjacent actual running times within the stagnation time period, and the angle of the vector formed by the actual location points corresponding to adjacent actual running times, a target discrimination index value is obtained for the stagnation time period. Next, based on the target discrimination index value, the DTW distance between the cumulative path length sequence corresponding to the stagnation time period in each actual transportation sub-time period and the cumulative path length sequence corresponding to the matching planned transportation path segment of the corresponding actual transportation sub-time period is adjusted to obtain the difference assessment result between the planned transportation path and the actual transportation path for the target anesthetic drug transportation task. Furthermore, this embodiment adjusts the DTW distance between the actual transportation path and the planned transportation path based on the target discrimination index value, which can improve the reliability of the similarity or difference assessment between the actual transportation path and the planned transportation path of the anesthetic drug transportation robot, thereby improving the accuracy of identifying and judging violations or risky behaviors such as route deviation, abnormal detours, and abnormal stops during actual transportation.

[0057] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for processing operational data of a robot for transporting anesthetic drugs, characterized in that, The method includes the following steps: The matching planned transportation path segments for each actual transportation sub-time period in the actual transportation time period corresponding to the target anesthetic drug transportation task are obtained, as well as the actual location points corresponding to each actual transportation time in each actual transportation sub-time period and the cumulative path length of each actual operation time, the planned location points on the matching planned transportation path segments, the cumulative path length of each planned location point, and the fixed stop planned location points corresponding to the target anesthetic drug transportation task. The stagnation time period in each actual transportation sub-time period is obtained based on the cumulative path length of the actual running time. The target discrimination index value corresponding to the stagnation time period is obtained based on the robot log of each actual running time in the stagnation time period, the distance between the actual position point corresponding to the actual running time in the stagnation time period and the fixed stop planning position point, the difference in cumulative path length between adjacent actual running times in the stagnation time period, and the angle of the vector formed by the actual position points corresponding to adjacent actual running times. Based on the target discrimination index value, the DTW distance between the cumulative path length sequence corresponding to the stagnation time period in each actual transportation sub-time period and the cumulative path length sequence corresponding to the matching planned transportation path segment of the corresponding actual transportation sub-time period is adjusted to obtain the difference evaluation result between the planned transportation path and the actual transportation path of the target anesthetic drug transportation task. The method for obtaining the cumulative path length includes: recording the path length from the starting point on the planned transportation path of the target anesthetic drug transportation task to each planned location point on the planned transportation path as the cumulative path length of the corresponding planned location point; obtaining the planned location point on the planned transportation path that is closest to the actual location point corresponding to each actual operating time, and recording it as the closest planned location point corresponding to the actual operating time; and using the cumulative path length of the closest planned location point corresponding to each actual operating time as the cumulative path length of the actual operating time. The method for obtaining the stagnation time periods in each actual transportation sub-time period includes: for any actual transportation sub-time period, obtaining feature points corresponding to each actual running time in the actual transportation sub-time period, where the horizontal coordinate of the feature point corresponding to any actual running time is the corresponding actual running time and the vertical coordinate is the cumulative path length of the corresponding actual running time; obtaining a slope feature value sequence corresponding to the actual transportation sub-time period based on the absolute value of the slope between the feature points corresponding to adjacent actual running times in the actual transportation sub-time period, where the v-th slope feature value in the slope feature value sequence is the absolute value of the slope between the feature points corresponding to the v-th and v+1-th actual running times in the actual transportation sub-time period; recording the result of negative correlation mapping after subtracting 0 from each slope feature value in the slope feature value sequence as the stagnation confidence characterization value corresponding to the slope feature value; and recording the time periods corresponding to consecutive slope feature values ​​in the slope feature value sequence whose stagnation confidence characterization values ​​are greater than a preset confidence threshold as stagnation time periods in the actual transportation sub-time period. The method for obtaining the target discrimination index value corresponding to the stagnation period includes: Obtain the normal business events corresponding to the target anesthetic drug delivery task, and based on the normal business events, the robot logs at each actual running moment in the stagnation period, and the distance between the actual location point corresponding to the actual running moment in the stagnation period and the fixed stop planning location point, obtain the first discriminant index value of the stagnation period; obtain the vector angle sequence corresponding to the stagnation period, where the f-th vector angle in the vector angle sequence refers to the angle between the actual location point corresponding to the f-th actual running moment in the stagnation period and the actual location point corresponding to the (f+1)-th actual running moment in the stagnation period, and based on the cumulative path length difference between adjacent actual running moments in the stagnation period and the vector angle sequence, obtain the second discriminant index value of the stagnation period; record the average of the first discriminant index value and the second discriminant index value of the stagnation period as the target discriminant index value corresponding to the stagnation period; The method for obtaining the evaluation results of the difference between the planned and actual delivery routes for the target anesthetic drug delivery task includes: All time periods during which the target discrimination index value is greater than the preset discrimination threshold are recorded as time periods to be adjusted. For any actual transportation sub-time period, all time periods in the actual transportation sub-time period except for the time period to be adjusted are recorded as non-adjusted time periods. The sequence formed by the cumulative path lengths of all planned location points on the matched planned transportation path segment of the actual transportation sub-time period is recorded as the matching sequence corresponding to the actual transportation sub-time period. The result of multiplying the DTW distance between the actual sequence corresponding to the time period to be adjusted and the matching sequence corresponding to the actual transportation sub-time period by the adjustment degree value of the corresponding time period to be adjusted is recorded as the target distance of the corresponding time period to be adjusted. The actual sequence corresponding to any time period is the sequence formed by the cumulative path lengths of all actual transportation moments in the corresponding time period. The adjustment degree value of any time period to be adjusted is the negative correlation mapping result of the target discrimination index value corresponding to the time period to be adjusted. The DTW distance between the actual sequence corresponding to the non-adjusted time period in the actual transportation sub-time period and the matching sequence corresponding to the actual transportation sub-time period is recorded as the target distance of the corresponding non-adjusted time period. The sum of the target distances of all non-adjusted time periods in the actual transportation sub-time period and the target distances of all time periods to be adjusted is recorded as the target distance of the actual transportation sub-time period. The sum of the target distances for all actual transport sub-time periods within the actual transport time period is recorded as the difference assessment result between the planned transport route and the actual transport route for the target anesthetic drug transport task.

2. The data processing method for a narcotic drug delivery robot as described in claim 1, characterized in that, The method for obtaining the matching and planned transportation route segments for each actual transportation sub-time period includes: The operational procedures stipulated or required when planning the delivery of narcotic drugs shall be recorded as necessary operational procedures for the delivery of narcotic drugs. The location on the planned transportation route of the target anesthetic drug transportation task that triggers the necessary business operation is recorded as the necessary location point on the planned transportation route. The moment when the necessary business operation is triggered in the actual transportation time period is recorded as the necessary moment. The actual transportation time period is segmented using the necessary moments in the actual transportation time period to obtain each actual transportation sub-time period in the actual transportation time period. For any actual transportation sub-time period, the two necessary moments in the actual transportation sub-time period are respectively recorded as the first moment and the second moment. The necessary location points that are the same as the necessary business operations triggered by the first moment are all recorded as the necessary location points corresponding to the first moment. Among the necessary location points corresponding to the first moment, the necessary location point closest to the actual location point corresponding to the first moment is obtained and recorded as the matching location point of the first moment. The method for obtaining the matching location point of the second moment is the same as the method for obtaining the matching location point of the first moment. The planned transportation path segment from the matching location point of the first moment to the matching location point of the second moment is taken as the matching planned transportation path segment of the actual transportation sub-time period.

3. The method for processing operational data of a narcotic drug delivery robot as described in claim 1, characterized in that, The method for obtaining the first discriminant index value of the stagnation period includes: Based on the robot logs and normal business events at each actual running moment during the stagnation period, feature label values ​​corresponding to each actual running moment during the stagnation period are obtained. If the robot log at any actual running moment contains the normal business event, then 1 is used as the feature label value corresponding to the corresponding actual running moment; otherwise, 0 is used as the feature label value corresponding to the corresponding actual running moment. The normalized result of the sum of the feature label values ​​corresponding to all actual running moments during the stagnation period is recorded as the first characterization value of the corresponding stagnation period. Calculate the average of the actual location coordinates corresponding to all actual running times during the stagnation period, and record it as the representative location point of the corresponding stagnation period. Calculate the negative correlation mapping result between the representative location point of the stagnation period and the nearest fixed stop planning location point to the representative location point of the stagnation period, and record it as the second characterization value of the corresponding stagnation period. Record the average of the first characterization value of the stagnation period and the second characterization value of the corresponding stagnation period as the first discrimination index value of the stagnation period.

4. The data processing method for a narcotic drug delivery robot as described in claim 1, characterized in that, The method for obtaining the second discriminant index value of the stagnation period includes: The set of absolute differences between the cumulative path lengths of all adjacent actual running times within the stagnation period is denoted as the cumulative path length difference sequence corresponding to the stagnation period. The result of summing all data in the cumulative path length difference sequence and then performing negative correlation mapping is denoted as the first mapping value of the stagnation period. The set of absolute differences between adjacent vector angles in the vector angle sequence corresponding to the stagnation period is denoted as the angle difference sequence corresponding to the stagnation period. The result of summing all data in the angle difference sequence and then performing negative correlation mapping is denoted as the second mapping value of the stagnation period. The mean of the first mapping value and the second mapping value of the stagnation period is denoted as the second discriminant index value of the stagnation period.

5. A data processing system for a robot transporting anesthetic drugs, comprising a memory and a processor, characterized in that, The processor executes the computer program stored in the memory to implement the business data processing method for a narcotic drug delivery robot as described in any one of claims 1-4.

Images