A console lifting control method based on artificial intelligence

Abstract

The application discloses a console lifting control method based on artificial intelligence, and relates to the technical field of intelligent man-machine interaction and mechatronic integration control.The application binds user parameters in real time through identity recognition, combines a height freezing mechanism in a continuous operation scene, and guarantees that new user operation cannot trigger sudden rising and sudden falling of a table top when a previous user has not logged out, thereby avoiding hidden troubles of misoperation in emergency dispatch.For unregistered temporary users, such as conference visitors, the application quickly estimates height through a full-body image and binds a temporary session identifier, provides seamless height adaptation during a session, and eliminates operation interruption problems of temporary users in a background that depend on manual adjustment.Combining with voiceprint feature recognition, the application guarantees that a voice instruction can trigger height recovery and improves cross-scene flexibility.A height adjustment priority is dynamically allocated based on arbitration decision of an operation time sequence, and when multiple people are intensively connected, such as customer service representatives who change shifts, the application automatically sorts execution order according to factors such as user roles and access frequency.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent human-computer interaction and mechatronic control, and particularly relates to a console lifting control method based on artificial intelligence. BACKGROUND

[0002] In multi-person cooperation scenes such as conference scheduling, customer service, and command center, the console often needs to be operated by users of different heights in turn. For example, in the daily cross-time zone video conference of a multinational enterprise, multiple participants need to quickly alternate speaking. When the shift of public service hotline operators is carried out, the operators are frequently replaced. Such scenes require the console lifting system to instantly adapt to the height difference of different users to avoid repeated manual adjustment affecting work efficiency.

[0003] The current mainstream scheme uses user identity recognition technology to realize height memory function, such as presetting height through face recognition or fingerprint binding. When a new user verifies identity, the system automatically calls the stored console position parameters of the user. However, in the multi-person continuous operation scene, this scheme has obvious limitations. If the user does not actively log out of the system after completing the operation, the subsequent user directly operates the console, and the device may follow the height setting of the previous user, causing the console to suddenly rise and fall. Especially in the emergency dispatching scene, the operator needs to quickly intervene in the device, and such sudden height switching is easy to cause misoperation risk. In addition, unregistered users, such as temporary visitors, cannot trigger the identity recognition process and still need to rely on physical button adjustment, which breaks the work continuity.

[0004] Some schemes introduce motion sensing mechanisms, such as detecting the user's departure through a pressure sensor to automatically reset the console height. However, in actual application, when multiple people are densely replaced, such as 3 operators being replaced within 15 minutes in a customer service, the sensor cannot distinguish between temporary absence and user replacement, and may misjudge as temporary absence and maintain the original height. Another type of scheme relies on a manual save key, which requires the user to actively confirm the setting after each operation, but this is easy to be ignored in high-load scenarios, which exacerbates operation confusion. Such methods do not solve the coordination contradiction between identity recognition and state switching, and there are still safety and efficiency risks when multiple people use it frequently. SUMMARY

[0005] In view of the above existing problems, the present application is proposed.

[0006] The present application provides a console lifting control method based on artificial intelligence to solve the problem that the existing scheme cannot coordinate the high-frequency rotation requirement due to the height mutation caused by the identity recognition not being logged out and the operation interruption of temporary users in the multi-person sharing scene.

[0007] To solve the above technical problems, the present application provides the following technical scheme:

[0008] The embodiment of the application provides a console lifting control method based on artificial intelligence, comprising the following steps:

[0009] In step S1, biological feature data close to a console operation area is collected in real time, and a user identity is generated;

[0010] In step S2, a user height parameter in a pre-stored database and an associated table height setting value are matched according to the identity;

[0011] In step S3, a target height setting value is called to drive a lifting mechanism to perform an action in response to a user state switching event;

[0012] In step S4, height arbitration decision is executed based on operation timing when a user identity change is detected in a continuous operation scenario.

[0013] As a preferred scheme of the console lifting control method based on artificial intelligence, the biological feature data collected in real time in step S1 comprises:

[0014] A user face image is captured through a console integrated camera;

[0015] A face image feature vector is extracted and compared with a registered database, and a unique identity is output.

[0016] As a preferred scheme of the console lifting control method based on artificial intelligence, the trigger condition of the user state switching event in step S3 comprises:

[0017] It is identified that a user limb contacts a console operation interface for more than a first time threshold value;

[0018] It is detected that seat pressure is zero and environmental audio input energy is lower than a silence threshold value for more than a second time threshold value.

[0019] As a preferred scheme of the console lifting control method based on artificial intelligence, in step S3, the determination mode of the first time threshold value T1 and the second time threshold value T2 comprises:

[0020] The duration distribution of intentional contact and accidental touch is usually bimodal, two normal curves are fitted according to historical labeled samples, and an operation threshold value is derived from the intersection point:

[0021]

[0022] Wherein, T1 represents the first time threshold value, the unit is s, sigma I is the standard deviation of intentional contact duration, the unit is s, mu N is the mean value of accidental contact duration, the unit is s, sigma NFor unintentional contact duration standard deviation, unit: s, μ I For intentional contact duration mean value, unit: s;

[0023] When the seat pressure is zero and the ambient sound energy enters the silent interval, the silent duration t s Also normally distributed; the upper limit of the missed detection probability P fn Calculate the quantile coefficient, and determine the second threshold value:

[0024] γ = Φ -1 (1-P fn ), T2 = μ S + γσ S ,

[0025] Where γ is the quantile coefficient, dimensionless, Φ -1 (·) is the inverse cumulative distribution function of the standard normal distribution, P fn is the allowed missed detection probability, 0-1, T2 is the second time threshold, unit: s, μ S ,σ S are the mean and standard deviation of the silent duration, respectively, unit: s;

[0026] An adaptive weight λ is introduced in the online phase, which automatically adjusts its size with contact stability:

[0027] Δσ = |σ I - σ N |,

[0028] Where λ is the recursive weight coefficient, 0-1, β is the sensitivity adjustment coefficient, and Δσ is the difference between the standard deviations of the two types of contact, unit: s;

[0029] Recursively update the threshold value using λ:

[0030]

[0031] Where, is the first-order threshold value after the kth update, s, is the threshold value of the last time, is the confirmed contact duration this time, and k is the online update serial number.

[0032] As a preferred scheme of the console lifting control method based on artificial intelligence, the height arbitration decision in step S4 includes:

[0033] Record the time when the previous user operation ends and the corresponding height setting value;

[0034] If the new user identity matching occurs within the third time threshold, freeze the current height setting value;

[0035] The freeze will be lifted and the height adjustment will be performed only upon receiving a voice confirmation command from a new user.

[0036] In a preferred embodiment of the artificial intelligence-based console lifting control method described in this invention, the determination of the continuous operation scenario includes:

[0037] Calculate the time interval between the start times of adjacent user operations;

[0038] When the time interval is less than the fourth time threshold, the height adjustment priority allocation is activated.

[0039] The execution order is sorted according to the frequency of user registration and access based on their identity.

[0040] As a preferred embodiment of the artificial intelligence-based console lifting control method of the present invention, wherein: in step S4, the calculation process of the fourth time threshold T4 includes:

[0041] In continuous operation scenarios, the handover time interval Δt between adjacent users follows a log-normal distribution; in the offline phase, the mean μ is obtained by fitting lnΔt. L With standard deviation σ L ; the upper limit of the probability of consecutive false rejection P c Under constraints, the fourth threshold is derived using quantiles:

[0042] φ=Φ -1 (1-P c ), T4 = exp(μ L -φσ L ),

[0043] Where φ is the quantile coefficient, which is dimensionless, Φ -1 (·) is the inverse cumulative distribution function of the standard normal distribution, P c To continuously determine the false rejection probability, a value of 0-1 is used, with T4 being the fourth time threshold, measured in seconds (μ). L σ is the mean of the logarithmic time intervals. L The standard deviation of the logarithmic time interval;

[0044] During the online phase, when the start time of the current user operation is detected... End time of the previous user operation The difference:

[0045] in, This indicates the start time of the operation triggered by the i-th user. Indicates the end time of the previous user's operation;

[0046] Satisfying Δt (i)When T4, it enters the height adjustment priority allocation mode;

[0047] Calculate the comprehensive priority score for the j-th user in the queue:

[0048]

[0049] where S j is the comprehensive priority score, dimensionless, j represents the ranking number of the user in the current waiting queue, w f , w r , w o are weight coefficients, dimensionless, satisfying w f + w r + w o = 1, f j is the historical access frequency of the user, in times per day, r j is the user role weight, 0 - 1, Δt j is the time interval between the user and the previous operation, s;

[0050] In the formula, the weights w f , w r , w o are adaptively adjusted according to the queue length N q :

[0051] w o = ε, w r = 1 - w f - w o ,

[0052] where ε is the minimum interval weight, k q is the queue sensitivity coefficient, in units of people -1 , N q is the current number of queuing users, in units of people, N0 is the inflection point queue length;

[0053] Execute the lifting action from high to low according to S j The arbitration process ends.

[0054] As a preferred solution of the console lifting control method based on artificial intelligence described in this invention, where: the following steps are executed for unregistered users:

[0055] Collect the full-body image of the user through the camera and estimate the height parameter based on the image ratio;

[0056] Generate a temporary session identifier and bind the estimated height parameter;

[0057] Automatically delete the temporary data after the session ends.

[0058] As a preferred scheme of the console lifting control method based on artificial intelligence, in the step S3, the lifting mechanism performs the action, which includes:

[0059] Outputting an acousto-optic early warning signal before height switching;

[0060] Monitoring the motor current change trend and dynamically adjusting the driving pulse width;

[0061] After reaching the target height, feeding back the position locking state to the control terminal.

[0062] As a preferred scheme of the console lifting control method based on artificial intelligence, in the step S3, the lifting mechanism performs the action, which includes:

[0063] Capturing the contact signal strength through the operation interface capacitive sensor;

[0064] When the signal strength continuously exceeds the reference level without interruption, starting the first time threshold timer.

[0065] The present application has the following advantages: the present application binds user parameters in real time through identity recognition, combines the height freezing mechanism in the continuous operation scene, and ensures that the new user operation will not trigger the sudden lifting and lowering of the console when the previous user has not logged out, avoiding the risk of misoperation in emergency scheduling; solve the defect of sudden lifting caused by the main user not logging out; for unregistered temporary users such as conference visitors, the height is quickly estimated through the full-body image and the temporary session identifier is bound, providing seamless height adaptation during the session, eliminating the operation interruption problem of temporary users in the background relying on manual adjustment; at the same time, combined with voiceprint feature recognition, the height recovery triggered by voice command is ensured, and the cross-scene flexibility is improved.

[0066] The present application dynamically allocates height adjustment priority based on arbitration decision of operation timing, and automatically sorts the execution order according to user roles, access frequency and other factors when multiple people are densely connected, such as customer service agent shift, avoiding the tedious process of repeated manual saving in traditional schemes, meeting the high-load demand of replacing multiple operators in a short clock; the lifting mechanism outputs an acousto-optic early warning before execution, giving the user response time to avoid; through motor current monitoring, the lifting speed is dynamically adjusted to prevent the console from tilting and overcome the risk of equipment impact caused by sudden leaving; the contact detection uses continuous capacitive signal strength judgment to exclude temporary false touch interference; the mute threshold is linked with the environmental audio energy to ensure the accuracy of leaving judgment; the two solve the reliability defect of sensor misjudgment of user state. BRIEF DESCRIPTION OF DRAWINGS

[0067] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description only constitute some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0068] Figure 1 Flowchart of the artificial intelligence-based console lifting control method in Embodiment 1. DETAILED DESCRIPTION

[0069] In order to make the above-mentioned purposes, features and advantages of the present application more apparent and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings of the specification.

[0070] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application. However, the present application can be practiced in other ways different from those described herein without departing from the scope of the present application, and those skilled in the art can make similar extensions without departing from the scope of the present application, so the present application is not limited to the specific embodiments disclosed below.

[0071] Secondly, the "one embodiment" or "embodiment" referred to herein means that the specific features, structures or characteristics can be included in at least one implementation of the present application. "In one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an independent or alternative embodiment that excludes other embodiments.

[0072] Embodiment 1, Reference Figure 1 The embodiment provides an artificial intelligence-based console lifting control method, comprising the following steps:

[0073] Step S1, real-time collection of biometric data close to the console operation area to generate a user identity;

[0074] The biometric data collected in real time in step S1 includes:

[0075] Capture the user's face image through the console integrated camera;

[0076] Extract the face image feature vector and compare it with the registration database to output a unique identity;

[0077] Step S2, according to the identity, match the user height parameter in the pre-stored database and the associated table height setting value;

[0078] Step S3, in response to a user state switching event, call the target height setting value to drive the lifting mechanism to execute the action;

[0079] The trigger conditions of the user state switching event in step S3 include:

[0080] The user's limb contact with the console operation interface is identified to last more than a first time threshold value;

[0081] The seat pressure is detected to be zeroed and the environmental audio input energy is lower than a silence threshold value for more than a second time threshold value;

[0082] In step S3, the determination of the first time threshold value T1 and the second time threshold value T2 includes:

[0083] The duration distribution of human contact and accidental touch is usually bimodal, and two normal curves are fitted according to historical labeled samples, and the operation threshold value is derived from the intersection point:

[0084]

[0085] Wherein, T1 represents the first time threshold value, the unit is s, σ I is the standard deviation of intentional contact duration, the unit is s, μ N is the mean value of unintentional contact duration, the unit is s, σ N is the standard deviation of unintentional contact duration, the unit is s, μ I is the mean value of intentional contact duration, the unit is s;

[0086] When the seat pressure is zeroed and the environmental sound energy enters the silence interval, the silence duration t s is also normally distributed; the upper limit of the missed detection probability P fn is calculated, and the second threshold value is determined:

[0087] γ=Φ -1 (1-P fn ), T2=μ S +γσ S ,

[0088] Wherein, γ is the quantile coefficient, dimensionless, Φ -1 (·) is the inverse cumulative distribution function of the standard normal distribution, P fn is the allowed missed detection probability, 0-1, T2 is the second time threshold value, the unit is s, μ S , σ S are the mean and standard deviation of silence duration, respectively, the unit is s;

[0089] An adaptive weight λ is introduced in the online stage, and its size is automatically adjusted according to the contact stability:

[0090] Δσ=|σ I -σ N |,

[0091] Wherein, λ is the recursive weight coefficient, 0-1, β is the sensitivity adjustment coefficient, 0.1-0.5s -1 , Δσ is the standard deviation difference of two types of contact, unit: s;

[0092] Recursively update the threshold value using λ:

[0093]

[0094] Wherein, is the first-order threshold value after the kth update, s, is the last threshold value, is the current confirmed contact duration, and k is the online update serial number;

[0095] Specifically, the intersection method makes the first threshold close to the place where the probabilities of two types of contact are equal, and effectively reduces the false touch rate in offline benchmark data; the second threshold uses quantile control to stably reduce the proportion of false positives in the silent period; the recursive module automatically converges according to contact stability: when the user's operation rhythm tends to be stable, Δσ decreases, λ decreases automatically, the threshold value changes with a low amplitude, the system output is smooth, when sudden acceleration or new users join, Δσ increases, λ increases, and the threshold value can be repositioned within a few interactions;

[0096] The lifting mechanism executes the action in step S3, which includes:

[0097] Output an audible and visual warning signal before switching the height;

[0098] Monitor the motor current trend and dynamically adjust the driving pulse width;

[0099] After reaching the target height, feed back the position locking state to the control terminal;

[0100] Identifying limb contact includes:

[0101] Capture the contact signal strength through the operation interface capacitive sensor;

[0102] When the signal strength continuously exceeds the reference level without interruption, start the first time threshold timer;

[0103] Step S4, when the user identity change is detected in the continuous operation scenario, execute height arbitration decision based on operation timing;

[0104] The height arbitration decision in step S4 includes:

[0105] Record the end time of the previous user operation and the corresponding height setting value;

[0106] If the new user identity match occurs within the third time threshold, freeze the current height setting value;

[0107] Unfreeze and perform height adjustment only after receiving a new user voice confirmation instruction;

[0108] The determination of the continuous operation scenario includes:

[0109] Calculate the time interval between the start times of adjacent user operations;

[0110] When the time interval is less than the fourth time threshold, activate the height adjustment priority allocation;

[0111] Sort the execution order according to the registered access frequency of the user identity identifier;

[0112] In step S4, the calculation process of the fourth time threshold T4 includes:

[0113] In the continuous operation scenario, the handover time interval Δt between adjacent users follows a lognormal distribution; the mean μ L and the standard deviation σ L are obtained by fitting lnΔt in the offline stage; under the constraint of the upper limit P c of the continuous determination false rejection probability, the fourth threshold is derived by quantile:

[0114] φ = Φ -1 (1 - P c ), T4 = exp(μ L - φσ L ),

[0115] where φ is the quantile coefficient, dimensionless, and Φ[[ID=​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​

[0121] wherein S j is the comprehensive priority score, dimensionless, j represents the ranking number of the user in the current waiting queue, w f ,w r ,w o is the weight coefficient, dimensionless, satisfying w f +w r +w o =1, f j is the user historical access frequency, unit: times / day, r j is the user role weight, 0-1, Δt j is the time interval between the user and the previous operation, s.

[0122] In the formula, the weight w f ,w r ,w o is adjusted adaptively with the queue length N q Self-adaptive adjustment:

[0123] w o =ε, w r =1-w f -w o ,

[0124] wherein ε is the minimum interval weight, k q is the queue sensitivity coefficient, unit: person -1 , N q is the current number of queued users, unit person, N0 is the inflection point queue length.

[0125] According to S j , the ascending and descending actions are executed from high to low, and the arbitration process ends.

[0126] Specifically, the fourth threshold is derived by logarithmic space quantile, so that the system can timely identify continuous scenarios when using intensively, and will not be triggered by mistake when using at low frequency; the adaptive weight highlights the role priority when the queue is short, gradually amplifies the influence of historical frequency and time interval when the queue increases, and maintains fairness.

[0127] The following steps are performed for unregistered users:

[0128] The user's full-body image is collected through the camera, and the height parameter is estimated based on the image scale;

[0129] A temporary session identifier is generated and bound to the estimated height parameter;

[0130] The temporary data is automatically deleted after the session is terminated.

[0131] The skilled in the art can calibrate or optimize the above parameters (such as β, ε, k q , etc.) according to specific application scenarios through routine experiments or statistical analysis methods, and thus the values thereof are not limited by the listed values in the embodiments of the present application.

[0132] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application but not limit the present application. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalently replaced without departing from the spirit and scope of the present application, and all modifications and equivalents should be included in the scope of the claims of the present application.

Claims

1. A method for controlling the lifting of a console based on artificial intelligence, characterized in that, Includes the following steps: Step S1: Collect biometric data in real time near the console operation area to generate user identity identifiers; Step S2: Match the user's height parameters and associated table height settings in the pre-stored database according to the identity identifier; Step S3: In response to the user state switching event, the target height setting value is invoked to drive the lifting mechanism to perform the action; Step S4: When a change in user identity is detected in a continuous operation scenario, a high-arbitration decision is made based on the operation sequence. The highly arbitrated decision-making process described in step S4 includes: Record the time when the previous user operation ended and the corresponding height setting value; If a new user's identity is matched within the third time threshold, freeze the current height setting. The freeze will be lifted and the height adjustment will be performed only upon receiving a voice confirmation command from a new user; The determination of the continuous operation scenario includes: Calculate the time interval between the start times of adjacent user operations; When the time interval is less than the fourth time threshold, the height adjustment priority allocation is activated. The execution order is sorted according to the frequency of user registration and access based on their identity.

2. The artificial intelligence-based console lifting control method as described in claim 1, characterized in that, The biometric data collected in real time in step S1 includes: Capture user facial images via integrated camera in the console; Extract facial image feature vectors and compare them with the registration database to output a unique identity identifier.

3. The artificial intelligence-based console lifting control method as described in claim 1, characterized in that, The triggering conditions for the user state switching event in step S3 include: The system detected that the user's body contact with the console interface exceeded the first time threshold. The system detected that the seat pressure had returned to zero and the ambient audio input energy was below the silence threshold for an extended period exceeding the second time threshold.

4. The artificial intelligence-based console lifting control method as described in claim 3, characterized in that, In step S3, the first time threshold With the second time threshold The methods for determining include: The duration distribution of human-induced and accidental touches often exhibits a bimodal characteristic. Based on historical labeled samples, two normal curves are fitted, and the operation threshold is derived from their intersection. ， in, This represents the first time threshold, in seconds. The standard deviation of the intentional contact duration, in seconds. This represents the average duration of unintentional contact, in seconds. The standard deviation of the duration of unintentional contact is expressed in seconds. The mean duration of intentional contact, in seconds; When the seat pressure reaches zero and the ambient sound energy enters the silence zone, the silence duration is... It also follows a normal distribution; with the upper limit of the false negative probability. Calculate the quantile coefficients, then determine the second threshold: ， ， in, These are quantile coefficients, dimensionless. It is the inverse cumulative distribution function of the standard normal distribution. To allow for the probability of missed detections, the value is between 0 and 1. This is the second time threshold, measured in seconds. These are the mean and standard deviation of the silent duration, respectively, in seconds. Introducing adaptive weights in the online phase Its size automatically adjusts based on contact stability: ， ， in, The recursive weight coefficient takes a value between 0 and 1. This is the sensitivity adjustment coefficient. The standard deviation difference between the two types of contact is expressed in seconds (s). use Recursively update the threshold: ， in, For the first The updated first-order threshold, in seconds. The threshold value from the previous time. This refers to the duration of the confirmed contact. To update the serial number online.

5. The artificial intelligence-based console lifting control method as described in claim 1, characterized in that, In step S4, the fourth time threshold The calculation process includes: In continuous operation scenarios, the handover time interval between adjacent users It exhibits a log-normal distribution; the offline stage is... The mean was obtained by fitting. with standard deviation ; Upper limit of the probability of consecutive false rejections Under constraints, the fourth threshold is derived using quantiles: ， ， in, These are quantile coefficients, dimensionless. It is the inverse cumulative distribution function of the standard normal distribution. To continuously determine the false rejection probability, the value ranges from 0 to 1. This is the fourth time threshold, measured in seconds. The mean of the logarithmic time intervals, The standard deviation of the logarithmic time interval; During the online phase, when the start time of the current user operation is detected... End time of the previous user operation The difference: ,in, Indicates the first The start time of the user-triggered operation. Indicates the end time of the previous user's operation; satisfy When this happens, it enters a high-priority allocation adjustment mode; For the first in the queue Calculate the overall priority score for each user: ， in, For comprehensive priority scoring, dimensionless. This indicates the rank number of the user currently in the waiting queue. The weighting coefficients are dimensionless and satisfy the following conditions: This represents the user's historical access frequency, expressed in times per day. This represents the user role weight, with a value between 0 and 1. The time interval between the user's previous action and the previous action, in seconds; In the formula, the weights With queue length Adaptive adjustment: ， ， ， in, The minimum interval weight, Queue sensitivity coefficient, unit: , This represents the current number of users in the queue, in units of people. The length of the inflection point queue; according to The arbitration process ends when the lifting and lowering motion is performed from high to low.

6. The artificial intelligence-based console lifting control method as described in claim 1, characterized in that, Perform the following steps for unregistered users: The user's full-body image is captured by a camera, and height parameters are estimated based on the image proportions. Generate a temporary session identifier and bind it to the estimated height parameter; Temporary data will be automatically deleted after the session ends.

7. The artificial intelligence-based console lifting control method as described in claim 1, characterized in that, The lifting mechanism in step S3 performs the following actions: Output audible and visual warning signals before altitude switching; Monitor the trend of motor current changes and dynamically adjust the drive pulse width; Once the target altitude is reached, the position lock status is reported back to the control terminal.

8. The artificial intelligence-based console lifting control method as described in claim 3, characterized in that, Identifying physical contact includes: The intensity of the contact signal is captured by a capacitive sensor through the user interface. When the signal strength continuously exceeds the reference level without interruption, the first time threshold timing is started.

Images