Intelligent scheduling method and device for coffee robot

Abstract

The invention provides an intelligent scheduling method and device for coffee robots, and the method comprises the steps: carrying out the real-time monitoring of the task state and resource availability of each coffee robot, and obtaining a comprehensive scheduling score of each coffee robot; when a new order is received, sorting the comprehensive scheduling scores to obtain a sorting list, and distributing the new order to the coffee robot with the highest comprehensive scheduling score in the sorting list; when monitoring that the monitoring data of a certain coffee robot in the coffee robots are abnormal, performing abnormal link judgment on the abnormal coffee robot, and after a corresponding problem link is obtained, querying the coffee robot with the highest score of the problem link of each monitoring data in the sorting list; and migrating the to-be-executed task or part of allocated tasks corresponding to the abnormal coffee robot to the coffee robot with the highest score in the problem link. By means of the scheme, the problems that in the prior art, scheduling precision is insufficient, exception handling is not timely, and task migration efficiency is not high are solved.

Description

Technical Field

[0001] This application relates to the technical field of intelligent scheduling, and in particular to an intelligent scheduling method and apparatus for a coffee robot. Background Technology

[0002] With the rapid development of the smart service industry, coffee robots are gradually becoming an important piece of equipment in commercial settings, replacing manual coffee making. They are characterized by high automation, stable production, and 24-hour operation, leading to their increasing deployment in cafes, office buildings, transportation hubs, and other locations. As the number of deployments increases, the stable operation of individual devices and the ability of multiple devices to collaboratively process orders are becoming key factors affecting operational efficiency and user experience.

[0003] In existing technologies, scheduling systems allocate new orders according to a preset order or based on the current number of tasks on each device, thereby achieving a roughly balanced workload among devices. Furthermore, when a device malfunctions or experiences an abnormal operation, the system switches tasks to other devices using fixed rules to ensure uninterrupted order production, thus enabling task assignment, simple load balancing, and basic failover capabilities.

[0004] Although a certain degree of task allocation and fault takeover can be achieved through fixed scheduling strategies or simple polling, problems still exist when the number of devices increases, the states of each coffee robot differ significantly, or there are multi-dimensional resource fluctuations. These problems include a single scheduling basis, incomplete resource assessment, crude anomaly judgment, and insufficient accuracy in task migration. Summary of the Invention

[0005] The technical problem to be solved by the present invention is the problem of insufficient scheduling accuracy, untimely anomaly handling and low task migration efficiency in the prior art, thereby providing an intelligent scheduling method and device for coffee robots.

[0006] In view of this, a first aspect of the present invention provides an intelligent scheduling method for coffee robots, comprising: real-time monitoring of the task status and resource availability of each coffee robot to obtain several monitoring data, and calculating a comprehensive scheduling score for each coffee robot based on the monitoring data; when a new order is received, sorting the comprehensive scheduling scores of each coffee robot to obtain a sorted list, and assigning the new order to the coffee robot with the highest comprehensive scheduling score in the sorted list; when an anomaly is detected in the monitoring data of one of the coffee robots, judging the abnormal link of the abnormal coffee robot to obtain the corresponding problem link; based on the problem link, querying the coffee robot with the highest problem link score in the sorted list, and migrating the pending task or part of the assigned task corresponding to the abnormal coffee robot to the coffee robot with the highest problem link score.

[0007] Optionally, the step of real-time monitoring of the task status and resource availability of each coffee robot to obtain several monitoring data points, and calculating the comprehensive scheduling score of each coffee robot based on the monitoring data, includes: acquiring monitoring data in real time through a data acquisition module deployed on each coffee robot; normalizing each parameter in the monitoring data and assigning corresponding weight coefficients to obtain multiple scoring sub-items; classifying each scoring sub-item to obtain a scoring sub-item representing the load level and a scoring sub-item representing the resource sufficiency; and weighting and summing the result of subtracting the score value of the scoring sub-item representing the load level from the preset maximum value and the score value representing the resource sufficiency to obtain the comprehensive scheduling score of each coffee robot.

[0008] Optionally, the monitoring data includes the number of orders in the current task queue, device CPU utilization, remaining completion time of the currently executing task, remaining coffee beans, remaining milk, remaining water tank volume, and device temperature parameters. The step of normalizing each parameter in the monitoring data and assigning it a corresponding weight coefficient to obtain multiple scoring sub-items, and classifying each scoring sub-item to obtain a scoring sub-item representing the load level and a scoring sub-item representing the resource sufficiency level, includes: normalizing the number of orders and assigning it a preset first weight coefficient to obtain a task load score; normalizing the device CPU utilization and assigning it a preset second weight coefficient to obtain a... The following steps are performed: 1. Prepare a load score; 2. Normalize the remaining completion time and assign a preset third weighting coefficient to obtain a time urgency score; 3. Normalize the remaining coffee beans, milk, and water tank volume, then weight and sum them, assigning a preset fourth weighting coefficient to obtain a resource sufficiency score; 4. Normalize the equipment temperature parameter and compare it with a preset normal temperature range, assigning a preset fifth weighting coefficient to obtain an equipment health score; 5. Construct a sub-item representing the load level based on the task load score, equipment load score, and time urgency score; 6. Construct a sub-item representing the resource sufficiency based on the resource sufficiency score and equipment health score.

[0009] Optionally, the step of sorting the comprehensive scheduling scores of each coffee robot to obtain a sorted list when a new order is received, and assigning the new order to the coffee robot with the highest comprehensive scheduling score in the sorted list, includes: parsing the new order to obtain order information when a new order is received; generating a sorted list in descending order of the comprehensive scheduling scores of each coffee robot at the current moment, and checking each coffee robot sequentially from the top of the sorted list to see if it meets the production capacity requirements of the new order; selecting the coffee robot in the sorted list that meets the production capacity requirements and has the highest comprehensive scheduling score as the optimal coffee robot; sending an order allocation instruction to the optimal coffee robot to write the order information of the new order to the end of the task queue of the optimal coffee robot, and updating the order quantity parameter in the monitoring data of the optimal coffee robot.

[0010] Optionally, the step of determining the abnormal link and corresponding problem link when the monitoring data of a coffee robot among the coffee robots is detected to be abnormal includes: when any monitoring data of a coffee robot exceeds a preset normal threshold range, the corresponding coffee robot is determined to be abnormal; the abnormal link of the abnormal coffee robot is determined, and if the CPU utilization rate of the abnormal coffee robot exceeds a preset first preset threshold or the temperature parameter of the device exceeds the normal temperature range, the calculation and processing link is marked as a problem link; if any of the remaining coffee beans, milk, or water tank of the abnormal coffee robot is lower than a preset second preset threshold, the raw material supply link is marked as a problem link; if the actual execution time of the currently executing task of the abnormal coffee robot exceeds a preset multiple of the estimated completion time, the production execution link is marked as a problem link; if the number of orders in the task queue of the abnormal coffee robot exceeds a preset third preset threshold and the duration exceeds a preset waiting time, the task scheduling link is marked as a problem link; when the judgment conditions of multiple problem links are met simultaneously, the main problem link is determined according to the preset problem link priority order.

[0011] Optionally, the step of querying the coffee robot with the highest score in the problem segment based on the problem segment in the sorted list, and migrating the pending tasks or some assigned tasks corresponding to the abnormal coffee robot to the coffee robot with the highest problem segment score, includes: determining the corresponding scoring dimension according to the type of the problem segment; filtering all normal coffee robots from the sorted list and extracting the parameter values ​​corresponding to the scoring dimension from the monitoring data of each normal coffee robot; normalizing the parameter values ​​and calculating the problem segment score of each coffee robot for the problem segment; selecting the coffee robot with the highest problem segment score as the task migration target device; extracting all orders in the pending execution state from the task queue of the abnormal coffee robot; if the progress of the task being executed by the abnormal coffee robot is less than a preset interruptible threshold, marking the task being executed as a migratable task; reordering the pending orders and the migratable tasks according to order priority and user waiting time, and inserting them into the task queue of the task migration target device according to the sorting result.

[0012] Optionally, the step of determining the corresponding scoring dimension based on the type of the problem stage includes: when the problem stage is a calculation and processing stage, the scoring dimension is the device CPU utilization rate and device temperature parameters; when the problem stage is a raw material supply stage, the scoring dimension is the remaining amount of coffee beans, milk, and water tank; when the problem stage is a production and execution stage, the scoring dimension is the remaining completion time of the currently executing task; and when the problem stage is a task scheduling stage, the scoring dimension is the number of orders in the task queue.

[0013] A second aspect of this invention provides an intelligent scheduling device for coffee robots, comprising: a monitoring module for real-time monitoring of the task status and resource availability of each coffee robot, obtaining several monitoring data, and calculating a comprehensive scheduling score for each coffee robot based on the monitoring data; an allocation module for sorting the comprehensive scheduling scores of each coffee robot when a new order is received, obtaining a sorted list, and allocating the new order to the coffee robot with the highest comprehensive scheduling score in the sorted list; a judgment module for judging the abnormal link of the coffee robot when an abnormality is detected in the monitoring data of one of the coffee robots, and obtaining the corresponding problem link; and a migration module for querying the coffee robot with the highest problem link score in the sorted list based on the problem link, and migrating the pending task or part of the assigned task corresponding to the abnormal coffee robot to the coffee robot with the highest problem link score.

[0014] The technical solution of the present invention has the following advantages: 1. By constructing a comprehensive scheduling score calculation mechanism based on monitoring data, the task status and resource availability of each coffee robot can be quantified in real time, thereby enabling dynamic evaluation and precise selection during task allocation.

[0015] 2. By introducing a structured monitoring data collection and processing flow, the comprehensive scheduling score can not only reflect the load of the coffee robot, but also the sufficiency of its raw material supply and the health of its hardware operation, making the scheduling results clear and interpretable. The sorting list built based on the comprehensive scheduling score makes the allocation process of new orders predictable and transparent, avoiding the repeated queuing and inefficient polling problems common in the traditional static scheduling mode, and improving the task response speed.

[0016] 3. By setting clear anomaly judgment thresholds and problem link classification mechanisms, specific abnormal links can be identified as soon as an anomaly occurs, thus providing an accurate basis for subsequent task migration steps. Furthermore, by setting corresponding scoring dimensions for problem links, the task migration process no longer relies on fixed rules, but is based on real-time calculations using monitoring data, thereby making the selection of task migration targets more reasonable and reliable.

[0017] This invention achieves efficient scheduling under normal conditions and rapid recovery under abnormal conditions, effectively ensuring the stability, continuity and efficiency of multiple coffee robots operating collaboratively, and improving the problems of insufficient scheduling accuracy, untimely abnormal handling and low task migration efficiency in the prior art. Attached Figure Description

[0018] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0019] Figure 1 A flowchart illustrating the intelligent scheduling method for a coffee robot provided in an embodiment of the present invention; Figure 2 This is a schematic block diagram of the intelligent scheduling device for a coffee robot provided in an embodiment of the present invention. Detailed Implementation

[0020] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] In the description of this invention, it should be noted that the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0022] Furthermore, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0023] like Figure 1 As shown in the example, this application provides an intelligent scheduling method for coffee robots. This intelligent scheduling method can evaluate the monitoring data of multiple coffee robots in real time and realize the dynamic allocation of orders and task migration in abnormal situations. The specific implementation includes the following steps: Step S1: Monitor the task status and resource availability of each coffee robot in real time to obtain a number of monitoring data, and calculate the comprehensive scheduling score of each coffee robot based on the monitoring data.

[0024] Data acquisition modules deployed on each coffee robot continuously collect monitoring data at a fixed sampling period (e.g., every 1 second). The monitoring data includes the number of orders in the task queue, device CPU utilization, remaining completion time of the currently executing task, remaining coffee beans, remaining milk, remaining water in the tank, and device temperature parameters. The monitoring data is saved to a central database and preprocessed based on preset data processing logic, including data format verification, parameter range verification, and timestamp correction. After preprocessing, a comprehensive scheduling score is calculated according to the evaluation rules corresponding to each parameter. Each evaluation rule has a clear evaluation direction and weight, ensuring that the comprehensive scheduling score accurately reflects the current load and resource sufficiency of the coffee robot.

[0025] For example, in a coffee robot A, the following data was collected: 4 orders in the task queue, 70% CPU utilization, 40 seconds remaining completion time, 30% remaining coffee beans, 45% remaining milk, 50% remaining water tank, and 65℃ equipment temperature. After normalizing these monitoring data, they were weighted and summed according to preset weights to obtain a comprehensive scheduling score of 0.62 for the coffee robot A, which reflects its current task load and resource availability.

[0026] Step S2: When a new order is received, sort the overall scheduling scores of each coffee robot to obtain a sorted list, and assign the new order to the coffee robot with the highest overall scheduling score in the sorted list.

[0027] The system monitors the order management module. When a new order is recorded, it parses the order to obtain the order type, production complexity level, and user priority identifier. After parsing, it reads the overall scheduling score of all coffee robots and generates a sorted list in descending order of score. Then, starting from the top of the sorted list, it checks whether each coffee robot meets the order's production capacity requirements, including whether it has the corresponding production program for the order type, whether there are enough remaining ingredients to complete the order, and whether the current device status is ready to accept orders.

[0028] When a new order is for an iced latte and the production complexity is level 3, the coffee robot A, coffee robot big, and coffee robot C in the sorted list are checked in turn. Although coffee robot A has a high overall scheduling score, its milk balance is only 10%, which is insufficient to make an iced latte. Therefore, coffee robot A is skipped, and coffee robot big, which meets the production capacity requirements and has the highest overall scheduling score, is selected as the task execution device.

[0029] Step S3: When an anomaly is detected in the monitoring data of one of the coffee robots, the abnormal link of the coffee robot is judged to obtain the corresponding problem link.

[0030] Threshold judgments are performed on the monitoring data, and the thresholds are preset parameter ranges based on actual operating experience. When any parameter in the monitoring data exceeds the corresponding threshold range, the coffee robot is marked as having an abnormal state, and the abnormal link is judged according to the type of monitoring data item that exceeds the threshold. For example, if the equipment CPU utilization rate is ≥95% or the equipment temperature parameter is ≥85℃, the calculation and processing link is judged to be abnormal; if the remaining coffee beans are ≤10%, the remaining milk is ≤10%, or the remaining water tank is ≤15%, the raw material supply link is judged to be abnormal; if the execution time exceeds twice the estimated completion time, the production execution link is judged to be abnormal; if the number of orders in the task queue is ≥8 and continues for more than 60 seconds, the task scheduling link is judged to be abnormal.

[0031] If the CPU utilization of coffee robot B reaches 97% and the equipment temperature reaches 90℃, the calculation and processing stage will be marked as a problem stage. If the remaining coffee beans are also less than 10%, the calculation and processing stage will be designated as the main problem stage based on the priority of problem stages.

[0032] Step S4: Based on the problem link query and sort list, the coffee robot with the highest problem link score for each monitoring data point will be migrated to the coffee robot with the highest problem link score.

[0033] The scoring dimensions are determined based on the problem stage. For example, the calculation and processing stage corresponds to the equipment's CPU utilization and temperature parameters; the raw material supply stage corresponds to the remaining coffee beans, milk, and water tank levels; the production and execution stage corresponds to the remaining completion time; and the task scheduling stage corresponds to the number of orders. All coffee robots in normal condition are selected from the sorted list, and the parameter values ​​corresponding to the scoring dimensions in their monitoring data are extracted, normalized, and then the problem stage score is calculated. The coffee robot with the highest score is selected as the target device for task migration. Subsequently, tasks to be executed are extracted from the task queue of the abnormal coffee robots, and it is determined whether the currently executing task can be interrupted (e.g., execution progress ≤ 30%). Migrationable tasks and tasks to be executed are reordered according to order priority and user waiting time and inserted into the task queue of the target device for task migration. The monitoring data and order binding relationship are updated, and the abnormal device is set to maintenance status.

[0034] Assuming the problem area is the raw material supply stage, then Coffee Robot Large, Coffee Robot C, and Coffee Robot D from the ranking list are selected as candidate devices. If Coffee Robot Large has 80% remaining coffee beans, 70% remaining milk, and 75% remaining water in its tank, and its score for the problem area after normalization is 0.88 (the highest score), then Coffee Robot Large is selected as the task migration target, and the orders and interruptible tasks that have not yet been executed by the faulty device B are migrated to Coffee Robot Large.

[0035] In this embodiment, the system monitors the task status and resource availability of each coffee robot in real time to obtain continuously updated monitoring data, and calculates a comprehensive scheduling score for each coffee robot based on the monitoring data. When the system receives a new order, it sorts the comprehensive scheduling scores of each coffee robot to generate a sorting list, and assigns the new order to the coffee robot with the highest comprehensive scheduling score according to the sorting list, thus realizing dynamic task scheduling. During operation, if an anomaly is detected in the monitoring data of one of the coffee robots, the system judges the abnormal link based on the monitoring data corresponding to the anomaly and identifies the problem link. Subsequently, the system queries the sorting list for the problem link score corresponding to the monitoring data of each coffee robot, selects the coffee robot with the highest problem link score as the task migration target, and migrates the pending tasks or part of the assigned tasks corresponding to the abnormal coffee robot to the target coffee robot, thereby completing the anomaly handling and task takeover process.

[0036] The above technical solution enables real-time evaluation of the operational status of each coffee robot, improving the accuracy of the comprehensive scheduling score calculation. This allows new orders to be assigned to the coffee robot with the highest comprehensive scheduling score in the sorting list, thereby enhancing the rationality and efficiency of overall task allocation. Simultaneously, when anomalies are detected in the monitoring data, the problem area can be more accurately located through anomaly identification. Based on the problem area, the coffee robot with the highest score in the problem area is selected from the sorting list to perform task migration, making the anomaly handling process more timely and accurate. This avoids order backlogs, task interruptions, or excessively long user waiting times due to anomalies in a single coffee robot. Overall, this technical solution significantly improves the task processing efficiency, anomaly recovery capability, and operational stability of the coffee robot cluster, providing a more reliable intelligent scheduling mechanism for multi-device collaborative work scenarios and addressing the problems of insufficient scheduling accuracy, untimely anomaly handling, and low task migration efficiency in existing technologies.

[0037] In another example, step S1 can preferably be implemented as follows: The data acquisition modules deployed on each coffee robot acquire monitoring data in real time. The monitoring data includes the number of orders in the current task queue, the device's CPU utilization, the remaining completion time of the currently executing task, the remaining amount of coffee beans, milk, and water in the tank, as well as the device's temperature parameters.

[0038] By configuring a data acquisition module inside each coffee robot, the module collects monitoring data at a fixed sampling period, not exceeding 1 second. The data is obtained through the robot's internal sensor interface, including order quantity, CPU utilization, remaining completion time, remaining coffee beans, remaining milk, remaining water level, and equipment temperature. The acquired monitoring data is uploaded in real-time by the data acquisition module to a monitoring interface, ensuring continuous monitoring of the operational status of all coffee robots.

[0039] For example, the data acquisition module extracts the equipment CPU utilization rate from the robot's main control board at a fixed interval of 1 second, uses an internal timer to count the remaining completion time of the currently executing task, uses a hopper level detector to obtain the remaining amount of coffee beans, milk, and water in the tank, and uses a temperature sensor to obtain the equipment temperature parameters. All of these are transmitted through the internal communication bus to continuously receive real-time monitoring data.

[0040] After normalizing each parameter in the monitoring data, corresponding weight coefficients are assigned to obtain multiple scoring sub-items. These sub-items are then classified to obtain scoring sub-items that represent the load level and scoring sub-items that represent the resource sufficiency level.

[0041] The monitored data, including order quantity, equipment CPU utilization, remaining completion time, remaining coffee beans, remaining milk, remaining water tank volume, and equipment temperature, are linearly normalized according to their respective preset minimum and maximum values, converting each parameter into a normalized value between 0 and 1. Subsequently, corresponding weight coefficients are assigned to each normalized parameter, and multiple scoring sub-items are generated based on a preset weight table. These scoring sub-items are then categorized by attribute into two groups: one representing the load level and the other representing the resource sufficiency level.

[0042] The order quantity was normalized to a range of 0 to 20, the equipment CPU utilization rate was normalized to a range of 0% to 100%, and the remaining coffee beans were normalized to a range of 0 to full load. Each of these was then multiplied by its respective weighting coefficient. The weighted order quantity score and CPU utilization rate score were then assigned to sub-items representing the load level, while the remaining coffee bean score and water tank score were assigned to sub-items representing the resource sufficiency level.

[0043] The comprehensive scheduling score for each coffee robot is obtained by subtracting the score of the sub-item representing the load level from the preset maximum value and then summing it with the score of the sub-item representing the resource sufficiency. The weight coefficients are dynamically updated based on historical scheduling data using a gradient descent algorithm.

[0044] The system calculates a comprehensive scheduling score for each coffee robot by subtracting the score of each sub-item representing the load level from a preset maximum value, ensuring that a higher score indicates a lighter load. This reverse processing is then performed, and the resulting score is weighted and summed with all sub-items representing resource sufficiency according to their respective weight coefficients. Historical scheduling data is used to record the actual impact of each weight coefficient on the scheduling effect, and the weights are adjusted using a gradient descent algorithm to make the comprehensive scheduling score more consistent with real-world device performance.

[0045] If the device's CPU utilization score is 0.7, and the preset maximum value is 1, then the reverse score is 1 - 0.7 = 0.3. The reversed CPU score and resource sufficiency score are multiplied by their corresponding weights of 0.3 and 0.5 respectively, and then summed to obtain the comprehensive scheduling score. The gradient descent algorithm uses the batch gradient descent method from existing technologies, making minor adjustments to the weights based on scheduling accuracy to make the next round of scoring results closer to the actual operating state.

[0046] Furthermore, the step of normalizing each parameter in the monitoring data and assigning corresponding weight coefficients to obtain multiple scoring sub-items, and classifying each scoring sub-item to obtain scoring sub-items representing the load level and scoring sub-items representing the resource sufficiency level, can be preferably as follows: After normalizing the order quantity, a preset first weight coefficient is assigned to obtain the task load score.

[0047] The order quantity is linearly normalized according to the preset order quantity upper limit, and the normalization result is multiplied by the first weight coefficient to make the impact of the order quantity on the task load score controllable and quantitative.

[0048] If the number of orders is 10 and the maximum number of orders is 20, the normalization result is 10 / 20=0.5. Multiplying 0.5 by the first weight coefficient of 0.2, we get a task load score of 0.1.

[0049] After normalizing the device's CPU utilization, a preset second weighting coefficient is assigned to obtain the device load score.

[0050] The device CPU utilization rate is normalized to a range of 0 to 100%, and the normalization result is multiplied by a second weighting coefficient to ensure that the weight of the device CPU utilization rate in the overall score meets the preset strategy.

[0051] If the device's CPU utilization is 80%, the normalized value is 0.8, and the second weighting coefficient is 0.3, then the device load score is 0.24.

[0052] The remaining completion time is normalized and then assigned a preset third weighting coefficient to obtain a time urgency score.

[0053] The remaining completion time is normalized according to the preset maximum acceptable completion time and multiplied by a third weighting coefficient, so that the urgency of the time is reflected in the final score.

[0054] If the remaining completion time is 30 seconds and the maximum acceptable completion time is 60 seconds, then the normalized value is 0.5, the third weighting coefficient is 0.2, and the time urgency score is 0.1.

[0055] The remaining coffee beans, milk, and water tank quantities are normalized, weighted, summed, and assigned a preset fourth weight coefficient to obtain a resource sufficiency score.

[0056] The remaining coffee beans, milk, and water tank quantities are normalized separately, then weighted and summed, and finally multiplied by a fourth weighting coefficient, so that the resource availability is reflected by a single score.

[0057] If the normalized result of the remaining coffee beans is 0.7, the normalized result of the remaining milk is 0.8, and the normalized result of the remaining water tank is 0.9, the sum of the three weighted by the ratios of 0.3, 0.3, and 0.4 is 0.82. Then, multiplying by the fourth weighting coefficient of 0.25, we get a resource sufficiency score of 0.205.

[0058] After normalizing the equipment temperature parameters, they are compared with the preset normal temperature range and then assigned a preset fifth weighting coefficient to obtain the equipment health score.

[0059] The equipment temperature parameters are compared with the upper and lower limits of the normal temperature range, and a health score is generated based on the deviation and multiplied by the fifth weighting coefficient.

[0060] If the equipment temperature parameter is 65℃, and the normal temperature range is 55℃ to 70℃, then the temperature is in the middle of the normal range, which is mapped to a normalized score of 0.6. Multiplying this by the fifth weighting coefficient of 0.15, we get an equipment health score of 0.09.

[0061] A scoring sub-item representing the degree of load is constructed based on task load score, equipment load score, and time urgency score.

[0062] The task load score, equipment load score, and time urgency score are combined into a load level score group to indicate the pressure on the equipment currently performing tasks.

[0063] If the three scores are 0.1, 0.24, and 0.1 respectively, then the sub-item representing the load level consists of these three scores, which are used for subsequent reverse processing.

[0064] Based on resource adequacy scores and equipment health scores, we construct scoring sub-items that characterize the degree of resource adequacy.

[0065] The resource sufficiency score and the equipment health score are combined to form a resource sufficiency score group, which is used to indicate the equipment's ability to support task execution.

[0066] If the resource adequacy score is 0.205 and the equipment health score is 0.09, then the two constitute the resource adequacy score sub-item, which is used for the final comprehensive score.

[0067] In another example, step S2 can preferably be implemented as follows: When a new order is received, it is parsed to obtain order information, which includes order type, production complexity level, and user priority identifier.

[0068] The system obtains the raw order data for new orders through the data receiving interface of the access layer. It then performs structured parsing on the raw order data, mapping the fields according to preset field formats for order type, production complexity level, and user priority identifier, and generates an order information object. The order type indicates the beverage category required for the new order; the production complexity level is identified based on the number of steps in the beverage preparation process, equipment usage time, and the number of types of raw materials involved; the user priority identifier is provided by the order source system when the order is placed and indicates the urgency or service level of the order. After parsing, the order information object immediately enters the order buffer queue to await allocation.

[0069] When receiving raw order data containing the fields {drink:“latte”, steps:7, materials:3, priority:“VIP”}, the field “latte” is mapped to the order type, the number of steps 7 and the number of material types 3 are converted to the production complexity level 3, “VIP” is mapped to the user priority identifier level 1, and finally an order information object {order type:latte, production complexity level:3, user priority identifier:1} is generated.

[0070] The overall scheduling scores of each coffee robot at the current moment are sorted in descending order to generate a sorted list. Starting from the top of the sorted list, each coffee robot is checked in turn to see if it meets the production capacity requirements for new orders. The production capacity requirements include whether it has the production program corresponding to the order type, whether there are enough remaining raw materials to complete the order, and whether the current equipment status is order-accepting.

[0071] A sorted list is generated based on the overall scheduling scores of each coffee robot at the current moment, from highest to lowest. The production capacity check program is then called sequentially based on the order of the sorted list. The production capacity check program reads the monitoring data of the coffee robots to be checked from the sorted list and retrieves whether the coffee robot has the production program corresponding to the order type from the coffee robot's capability description file. Simultaneously, it compares the required raw material quantity in the order information with the remaining coffee beans, milk, and water tank quantity in the monitoring data to determine if the remaining raw materials are sufficient to complete the order. Finally, it checks the equipment status field in the monitoring data to confirm whether its status is "order-acceptable." Only when all three conditions are met is the coffee robot deemed to meet the production capacity requirements.

[0072] For example, if a coffee robot A in the sorted list has a comprehensive scheduling score of 85, and the monitoring data shows that the device has a latte-making program, with 45g of coffee beans, 300ml of milk, and 350ml of water remaining, all of which are higher than the raw material thresholds required for latte making (20g of coffee beans, 150ml of milk, and 200ml of water), and the device status field is "idle" (order-accepting status), then it is determined that coffee robot A meets the production capacity requirements.

[0073] The coffee robot that meets the production capacity requirements and has the highest overall scheduling score in the sorted list is selected as the optimal coffee robot.

[0074] Starting from the top of the sorted list, each coffee robot is checked one by one to ensure it meets the production capacity requirements, and the first coffee robot to meet the requirements is recorded. Based on the comprehensive scheduling score sorting logic, the system ensures that the first coffee robot to meet the production capacity requirements is the one with the highest comprehensive scheduling score. This coffee robot will be selected as the optimal coffee robot and locked as the target device for this task.

[0075] The sorted list, ranked from highest to lowest score, is A(85), B(79), and C(72). If A meets the production capacity requirements, then A becomes the optimal coffee robot; if A does not meet the production capacity requirements but B does, then B is selected as the optimal coffee robot.

[0076] Send an order allocation instruction to the optimal coffee robot to write the order information of the new order to the end of the task queue of the optimal coffee robot, update the order quantity parameter in the monitoring data of the optimal coffee robot, and record the binding relationship between the order and the target allocation device and the allocation timestamp in the order management database.

[0077] An order allocation instruction is sent to the optimal coffee robot. This instruction includes order information, the target task queue position, the task identifier, and the execution timestamp. Upon receiving the instruction, the optimal coffee robot writes the order information to the end of its task queue, explicitly placing the new order in the pending execution list. Simultaneously, the order quantity parameter in the optimal coffee robot's monitoring data is updated by incrementing the parameter by 1. The binding relationships between the order number, optimal coffee robot number, allocation timestamp, and user priority identifier are recorded in the order management database to ensure subsequent task tracking, migration, or scheduling backtracking.

[0078] If the order number is O23112 and the new order is assigned to coffee robot B, the record {Order Number: O23112, Bound Device: B, Assignment Time: 2025-11-18 14:12:08, User Priority Identifier: 1} will be written, and the order quantity parameter of B will be updated from 3 to 4.

[0079] In another example, step S3 can preferably be implemented as follows: When any monitoring data of a coffee robot exceeds the preset normal threshold range, the corresponding coffee robot is determined to be abnormal.

[0080] The monitoring data of each coffee robot is periodically read and compared field by field with the corresponding normal threshold range. When any monitoring data field exceeds the corresponding normal threshold, the coffee robot is immediately marked as abnormal, and the abnormal field is recorded in the abnormal mark list as the basis for subsequent abnormal judgment.

[0081] If the normal threshold for the equipment temperature parameter of a coffee robot is below 45℃, but the monitoring data shows that the temperature has reached 58℃, the system will immediately determine that the coffee robot is abnormal and record the abnormal field "equipment temperature parameter".

[0082] For coffee robots that malfunction, an abnormal process is identified. If the CPU utilization rate of the malfunctioning coffee robot exceeds the first preset threshold or the temperature parameter of the device exceeds the normal temperature range, the calculation and processing process is marked as a problem process.

[0083] Check the list of anomaly markers. If the list contains fields for device CPU utilization or device temperature parameters, mark the computing and processing stage as a problem stage. If the device CPU utilization exceeds the first preset threshold, it indicates that the computing unit is overloaded. If the device temperature parameter exceeds the normal temperature range, it indicates abnormal heat dissipation or hardware overload. Both can be attributed to a fault in the computing and processing stage.

[0084] If the first preset threshold is 90%, and the CPU utilization monitoring data of a coffee robot shows 97%, then the calculation and processing stage will be marked as a problem stage.

[0085] If any of the coffee bean, milk, or water tank levels of the abnormal coffee robot falls below a preset second threshold, the raw material supply process will be marked as a problem process.

[0086] The system reads the abnormal fields related to raw materials from the abnormal flag list. If any of the remaining coffee beans, milk, or water tank levels is below the second preset threshold, the raw material supply process is marked as a problem process based on the category of the field. The second preset threshold corresponds to the minimum safe raw material level in the equipment operation manual.

[0087] If the second preset threshold is 10g of coffee beans, 80ml of milk, and 150ml of water, and the remaining milk monitoring data of a certain coffee robot shows only 50ml, the system will mark the raw material supply link as a problem link.

[0088] If the actual execution time of a coffee robot currently performing a task exceeds a preset multiple of the estimated completion time, the production execution stage will be marked as a problem stage.

[0089] The system reads the actual execution time field from the monitoring data of the malfunctioning coffee robot and compares it with the estimated completion time in the order information. If the actual execution time exceeds a preset multiple, such as more than twice the estimated completion time, it indicates that the production process is obstructed, such as internal mechanical structure jamming or interruption of the production process. The system then marks the production process as a problematic step.

[0090] If the estimated completion time for a latte is 45 seconds, but the actual execution time of a coffee robot has reached 95 seconds, the system will mark the production process as a problem.

[0091] If the number of orders in the task queue of an abnormal coffee robot exceeds the preset third threshold and the duration exceeds the preset waiting time, the task scheduling process will be marked as a problem process.

[0092] The system checks whether the number of orders in the task queue of the coffee robot experiencing the anomaly exceeds a third preset threshold, and whether the duration exceeds a preset waiting time. If both conditions are met, it indicates that task scheduling has failed to process orders in a timely manner or that there is an order backlog; therefore, the task scheduling process is marked as a problem.

[0093] If the third preset threshold is "the number of orders in the task queue is ≥10", the preset waiting time is 120 seconds; if the number of orders in the task queue of a coffee robot is 13, and this state has lasted for 168 seconds, then the system will mark the task scheduling process as a problem process.

[0094] When multiple problem conditions are met simultaneously, the primary problem is determined according to the preset priority order of the problem stages.

[0095] Multiple problem stages are sorted according to a preset priority order to determine the primary problem stage. The priority order is set based on the degree of impact on task execution; for example, the calculation and processing stage has a higher priority than the production and execution stage, the production and execution stage has a higher priority than the raw material supply stage, and the raw material supply stage has a higher priority than the task scheduling stage. The system selects the highest priority problem stage from the priority list as the primary problem stage.

[0096] If a coffee robot simultaneously triggers abnormal equipment temperature parameters (calculation and processing stage) and insufficient raw materials (raw material supply stage), the calculation and processing stage has higher priority according to the priority rules, and the system will treat the calculation and processing stage as the main problem stage.

[0097] In another example, step S4 can preferably be implemented as follows: The scoring dimensions are determined based on the type of problem stage. For example, if the problem stage is a computational processing stage, the scoring dimensions are the equipment's CPU utilization and temperature parameters; if the problem stage is a raw material supply stage, the scoring dimensions are the remaining coffee beans, milk, and water tank levels; if the problem stage is a production execution stage, the scoring dimension is the remaining completion time of the currently executing task; and if the problem stage is a task scheduling stage, the scoring dimension is the number of orders in the task queue. All coffee robots in normal condition are then selected from the sorted list, and the parameter values ​​corresponding to the scoring dimensions are extracted from the monitoring data of each normal coffee robot.

[0098] Based on the identified problem stage type, pre-defined scoring dimensions are matched, and all normally functioning coffee robots are selected from the sorted list. Monitoring data from these normally functioning robots is read, and parameter values ​​corresponding to the scoring dimensions are extracted. The scoring dimensions are mapped one-to-one with the problem stage. For example, the scoring dimensions for the calculation and processing stage are the device's CPU utilization and temperature parameters; for the raw material supply stage, they are the remaining coffee beans, milk, and water tank levels; for the production execution stage, the remaining completion time of the currently executing task; and for the task scheduling stage, the number of orders in the task queue. Simultaneously, the parameter values ​​are formatted and validated to ensure valid input values. A parameter value index table is created according to the coffee robot's ID to ensure accurate association with each normally functioning coffee robot in subsequent calculations.

[0099] When the problem area is the raw material supply stage, all coffee robots with a "normal" status are selected from the sorted list. The remaining coffee bean quantity, milk quantity, and water tank quantity for each coffee robot are extracted from the monitoring data. These parameter values ​​are then indexed according to "coffee robot number - corresponding raw material quantity". For example, coffee robot A has a coffee bean quantity of 70%, milk quantity of 65%, and water tank quantity of 80%, while coffee robot B has a coffee bean quantity of 85%, milk quantity of 75%, and water tank quantity of 90%. The index table will record the three types of parameter values ​​corresponding to A and B, which will be used for subsequent calculation of the problem area score.

[0100] After normalizing the parameter values, the problem segment scores of each coffee robot are obtained, and the coffee robot with the highest problem segment score is selected as the target device for task migration.

[0101] The parameter values ​​extracted from the monitoring data are normalized using the existing linear normalization method, which converts each parameter value into a score within the range [0,1] according to the formula "(parameter value - minimum value) / (maximum value - minimum value)". After normalization, the normalized scores are weighted and summed according to the scoring dimension. The weight coefficients are pre-set based on the influence of the scoring dimension in historical scheduling records to obtain the problem stage score for each coffee robot. Simultaneously, the coffee robot with the highest problem stage score among all normal coffee robots is selected as the target device for task migration.

[0102] When the problem stage is the computational processing stage, the scoring dimensions are the device's CPU utilization and device temperature parameters. Two score values ​​are obtained through linear normalization. For example, the normalized value of the device's CPU utilization is 0.3 and the normalized value of the device temperature parameter is 0.4. The scores are then weighted and summed according to preset weighting coefficients (e.g., the weight of the device's CPU utilization is 0.6 and the weight of the device temperature parameter is 0.4) to obtain a score of 0.34 for the computational processing stage. Similarly, the same calculation process is performed on all coffee robots in normal state, and the one with the highest score in the problem stage is selected as the target device for task migration.

[0103] Extract all orders in the pending execution state from the task queue of the coffee robot that malfunctioned. If the progress of the task being executed by the coffee robot that malfunctioned is less than the preset interruptible threshold, mark the task being executed as a transferable task.

[0104] The system reads all pending orders from the task queue of the malfunctioning coffee robot and determines whether the progress of the currently executing task is below a preset interruptibility threshold. This threshold is a system-defined percentage of task progress that can be interrupted, such as 30%. When the task progress falls below this threshold, the currently executing task is marked as a transferable task to ensure that interruption does not lead to a decrease in production quality or a degraded user experience. Simultaneously, pending orders and transferable tasks are recorded separately, and a task migration list is generated. This list includes task parameters such as order number, user information, order type, production complexity level, and remaining execution time, for subsequent task reordering and migration operations.

[0105] For example, the task queue of the coffee robot C that malfunctioned contains three orders in the pending execution state: order X, order Y, and order Z. At the same time, the task being executed by the coffee robot C has only completed 20% of its execution progress, which is lower than the preset interruptible threshold of 30%. Therefore, the task being executed and the three orders in the pending execution state are recorded together as migration tasks in the task migration list, and the task information is structured to ensure that all parameters of the migration task can be used for subsequent sorting and task insertion operations.

[0106] Orders awaiting execution and transferable tasks are reordered according to order priority and user waiting time, and then inserted into the task queue of the target device for task migration according to the sorting result. At the same time, the monitoring data of the target device for task migration is updated, and the device binding relationship of the relevant orders is updated in the order management database. The coffee robot with abnormal status is marked as maintenance status and the acceptance of new orders is suspended.

[0107] Orders and migrateable tasks in the task migration list are reordered based on order priority and user waiting time. The sorting method is as follows: first, sort by user priority identifier from high to low; when multiple orders have the same user priority identifier, then sort by user waiting time from long to short. After sorting, tasks are inserted into the task queue of the target device or a specified insertion position according to the sorted task list. After insertion, the monitoring data of the target device is updated synchronously with parameters such as order quantity, coffee bean balance, milk balance, water tank balance, and remaining completion time. At the same time, the order number, device binding relationship, and migration timestamp corresponding to the migration task are updated in the order management database. The device status of coffee robots with abnormalities is marked as maintenance status, suspending them from receiving new orders and preventing abnormal devices from participating in task scheduling again.

[0108] During the migration task sorting process, if the user priority of order X, which is pending execution, is higher than that of orders Y and Z, and the user waiting time for the currently executing task is the longest, then the sorting result is "currently executing task - order X - order Y - order Z". Subsequently, these four sorted tasks are inserted into the end of the task queue of the target device for task migration. For example, when the target device for task migration is coffee robot B, the four tasks are written into the task queue of coffee robot B. At the same time, the order quantity parameter in the monitoring data of coffee robot B is increased from the original quantity to the new order quantity, and the increased task execution load after inserting the tasks is calculated. The migration timestamp (e.g., November 18, 2025, 14:32:10) and the updated device binding relationship are also recorded in the order management database to ensure that the order flow process is completely traceable. Coffee robot C, which has an abnormality, is marked as maintenance status, and its subsequent task receiving function is suspended.

[0109] like Figure 2 As shown, this application also provides an intelligent scheduling device 10 for a coffee robot, including the following modules: Monitoring module 11 is mainly used to monitor the task status and resource availability of each coffee robot in real time, obtain a number of monitoring data, and calculate the comprehensive scheduling score of each coffee robot based on the monitoring data.

[0110] The monitoring module 11 continuously monitors the task status and resource availability of each coffee robot. It can periodically generate monitoring data during task execution and dynamically update the comprehensive scheduling score of each coffee robot based on the monitoring data. This allows the comprehensive scheduling score to reflect the execution capability, idle time, and resource load of each coffee robot in real time.

[0111] The allocation module 12 is mainly used to sort the comprehensive scheduling scores of each coffee robot when a new order is received, obtain a sorted list, and allocate the new order to the coffee robot with the highest comprehensive scheduling score in the sorted list.

[0112] After receiving a new order, the allocation module 12 can build a sorting list based on the latest comprehensive scheduling score and automatically allocate the new order to the coffee robot with the highest comprehensive scheduling score in the sorting list, making the task allocation process real-time and targeted.

[0113] Judgment module 13 is mainly used to judge the abnormal link of a coffee robot when the monitoring data of a coffee robot among the coffee robots is detected to be abnormal, and to obtain the corresponding problem link.

[0114] When the monitoring module 11 detects an anomaly in the monitoring data, the judgment module 13 can determine the abnormal link of the coffee robot and identify the specific problem link based on the type of monitoring data corresponding to the anomaly, the time of the anomaly trigger, and the duration of the anomaly.

[0115] Migration module 14 is mainly used to migrate the coffee robot with the highest score in each problem area in the problem area query and sort list to the coffee robot with the highest score in the problem area.

[0116] The migration module 14 is based on the problem stage. It queries the sorted list for the coffee robot with the highest score in the corresponding problem stage. This enables the system to complete the task migration when the target coffee robot has the advantage in dealing with the abnormal stage. The system can smoothly migrate the pending tasks or some of the assigned tasks corresponding to the abnormal coffee robot to the coffee robot, so as to ensure that the overall scheduling process can still maintain service continuity, task reachability and scheduling stability when a local abnormality occurs.

[0117] It should be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the device and each module described above can be referred to the corresponding process in the aforementioned Embodiment 1, and will not be repeated here.

[0118] The above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. An intelligent scheduling method for a coffee robot, characterized in that, include: The task status and resource availability of each coffee robot are monitored in real time to obtain a number of monitoring data. The comprehensive scheduling score of each coffee robot is calculated based on the monitoring data. When a new order is received, the overall scheduling scores of each coffee robot are sorted to obtain a sorted list, and the new order is assigned to the coffee robot with the highest overall scheduling score in the sorted list. When an anomaly is detected in the monitoring data of one of the coffee robots, the abnormal link of the coffee robot is identified to determine the corresponding problem link. Based on the problem stage, query the coffee robot with the highest problem stage score in the sorted list of monitoring data, and migrate the pending tasks or some assigned tasks corresponding to the abnormal coffee robot to the coffee robot with the highest problem stage score.

2. The intelligent scheduling method for coffee robots according to claim 1, characterized in that, The steps of real-time monitoring of the task status and resource availability of each coffee robot to obtain several monitoring data points, and calculating the comprehensive scheduling score of each coffee robot based on the monitoring data, include: The data acquisition module deployed on each coffee robot acquires monitoring data in real time. After normalizing each parameter in the monitoring data, it assigns a corresponding weight coefficient to obtain multiple scoring sub-items. Each scoring sub-item is classified to obtain a scoring sub-item that represents the load level and a scoring sub-item that represents the resource sufficiency level. The overall scheduling score for each coffee robot is obtained by subtracting the score of the sub-item representing the load level from the preset maximum value and then weighting and summing it with the score of the sub-item representing the resource sufficiency.

3. The intelligent scheduling method for coffee robots according to claim 2, characterized in that, The monitoring data includes the number of orders in the current task queue, the device's CPU utilization rate, the remaining completion time of the currently executing task, the remaining amount of coffee beans, milk, and water in the tank, as well as the device's temperature parameters. The step of normalizing each parameter in the monitoring data and assigning it a corresponding weight coefficient to obtain multiple scoring sub-items, classifying each scoring sub-item to obtain scoring sub-items representing the load level and scoring sub-items representing the resource sufficiency level, includes: After normalizing the order quantity, a preset first weighting coefficient is assigned to obtain the task load score. After normalizing the CPU utilization rate of the device, a preset second weighting coefficient is assigned to obtain the device load score. The remaining completion time is normalized and then assigned a preset third weighting coefficient to obtain a time urgency score. The remaining coffee beans, milk, and water tank quantities are normalized, weighted, and summed, and then assigned a preset fourth weight coefficient to obtain a resource sufficiency score. After normalizing the equipment temperature parameters, they are compared with a preset normal temperature range and then assigned a preset fifth weighting coefficient to obtain the equipment health score. Based on the task load score, the device load score, and the time urgency score, a scoring sub-item representing the load level is constructed; Based on the resource sufficiency score and the equipment health score, a scoring sub-item representing the degree of resource sufficiency is constructed.

4. The intelligent scheduling method for coffee robots according to claim 1, characterized in that, The step of sorting the overall scheduling scores of each coffee robot when a new order is received to obtain a sorted list, and assigning the new order to the coffee robot with the highest overall scheduling score in the sorted list, includes: When a new order is received, the new order is parsed to obtain the order information; The overall scheduling scores of each coffee robot at the current moment are sorted in descending order to generate a sorting list. Starting from the top of the sorting list, each coffee robot is checked in turn to see if it meets the production capacity requirements for new orders. The coffee robot that meets the production capacity requirements and has the highest comprehensive scheduling score in the sorted list is selected as the optimal coffee robot. Send an order allocation instruction to the optimal coffee robot to write the order information of the new order to the end of the task queue of the optimal coffee robot, and update the order quantity parameter in the monitoring data of the optimal coffee robot.

5. The intelligent scheduling method for coffee robots according to claim 3, characterized in that, The step of determining the abnormal process and identifying the corresponding problem process when an anomaly is detected in the monitoring data of one of the coffee robots includes: When any monitoring data of a coffee robot exceeds the preset normal threshold range, the corresponding coffee robot is determined to be abnormal. For coffee robots that malfunction, an abnormal process is identified. If the CPU utilization rate of the malfunctioning coffee robot exceeds a preset first threshold or the temperature parameter of the device exceeds the normal temperature range, the calculation and processing process is marked as a problem process. If any of the remaining coffee beans, milk, or water tank of the abnormal coffee robot is lower than a preset second preset threshold, the raw material supply link will be marked as a problem link. If the actual execution time of the currently performing task of the coffee robot that is malfunctioning exceeds a preset multiple of the estimated completion time, the production execution stage will be marked as a problem stage. If the number of orders in the task queue of the coffee robot that is malfunctioning exceeds a preset third preset threshold and the duration exceeds a preset waiting time, the task scheduling process will be marked as a problem process. When multiple problem conditions are met simultaneously, the primary problem is determined according to the preset priority order of the problem stages.

6. The intelligent scheduling method for a coffee robot according to claim 5, characterized in that, The step of querying the coffee robot with the highest problem score in the sorted list based on the problem problem, and migrating the pending tasks or part of the assigned tasks corresponding to the abnormal coffee robot to the coffee robot with the highest problem problem score, includes: Based on the type of the problem, determine the corresponding scoring dimension, filter out all coffee robots in normal condition from the sorting list, and extract the parameter values ​​corresponding to the scoring dimension from the monitoring data of each normal coffee robot. After normalizing the parameter values, the problem segment score of each coffee robot for the problem segment is obtained, and the coffee robot with the highest problem segment score is selected as the target device for task migration. Extract all orders in the pending execution state from the task queue of the coffee robot that malfunctioned. If the progress of the task being executed by the coffee robot that malfunctioned is less than the preset interruptible threshold, mark the task being executed as a transferable task. After reordering the orders in the pending execution state and the transferable tasks according to order priority and user waiting time, they are inserted into the task queue of the target device for task migration according to the sorting result.

7. The intelligent scheduling method for coffee robots according to claim 6, characterized in that, The step of determining the corresponding scoring dimension based on the type of the problem segment includes: When the problem stage is a computational processing stage, the scoring dimensions are the device CPU utilization and the device temperature parameter; When the problem area is the raw material supply area, the scoring dimensions are the remaining amount of coffee beans, milk, and water tank. When the problem stage is the production and execution stage, the scoring dimension is the remaining completion time of the currently executing task; When the problematic stage is the task scheduling stage, the scoring dimension is the number of orders in the task queue.

8. An intelligent scheduling device for a coffee robot, characterized in that, include: The monitoring module is used to monitor the task status and resource availability of each coffee robot in real time, obtain a number of monitoring data, and calculate the comprehensive scheduling score of each coffee robot based on the monitoring data. The allocation module is used to sort the comprehensive scheduling scores of each coffee robot when a new order is received, obtain a sorted list, and allocate the new order to the coffee robot with the highest comprehensive scheduling score in the sorted list. The judgment module is used to determine the abnormal link of the coffee robot when the monitoring data of one of the coffee robots is found to be abnormal. The migration module is used to query the coffee robot with the highest score in each monitoring data of the problem link in the sorted list based on the problem link, and migrate the pending tasks or some assigned tasks corresponding to the coffee robot with the abnormality to the coffee robot with the highest score in the problem link.

Images