A vertical circulating three-dimensional intelligent garage self-service car taking system
By acquiring multi-dimensional user status data, performing feature fusion and judgment, and triggering personalized scheduling modes, the problem of existing vertical circulation self-service parking garage systems being unable to recognize user status and lacking emotional interaction has been solved, thus improving service intelligence and fairness.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- GUIZHOU XINZHAO INTELLIGENT TECH CO LTD
- Filing Date
- 2026-02-24
- Publication Date
- 2026-05-29
AI Technical Summary
Existing vertical circulation parking garage self-service retrieval systems cannot recognize differences in user status, lack emotional interaction, have insufficient intelligent service capabilities, and cannot provide adaptive services when users face inconvenience.
The system employs a perception acquisition layer to acquire multi-dimensional state data, a feature processing layer to parse and extract core features, a state determination layer to perform multi-feature fusion to determine the user's state, a mode triggering layer to trigger the corresponding scheduling mode, a scheduling execution layer to execute vehicle scheduling, and a fair compensation layer to compensate affected users.
It enables reliable judgment of users' emotions and states, provides personalized services, enhances the intelligence and humanization of the system, ensures the fairness and security of scheduling, and reduces user disputes.
Smart Images
Figure CN122106310A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of intelligent parking garage technology, specifically relating to a self-service car retrieval system for a vertical circulation three-dimensional intelligent parking garage. Background Technology
[0002] Vertical circulation parking garages are widely used in densely populated urban areas due to their advantages such as small footprint and high space utilization. Existing self-service parking garage systems typically use a central control unit for centralized scheduling. Their operating mode is as follows: after the user inputs a retrieval command, the central controller calculates the movement sequence of the vehicle carrier platform based on a fixed algorithm (such as shortest path or sequential search), and then controls the motors to rotate forward and backward in sequence, circulating the target vehicle carrier platform to the exit position.
[0003] Existing self-service car retrieval systems in vertical circulation parking garages rely solely on the timing or fixed rules of requests, treating all user requests indiscriminately without recognizing objective differences in user states. The system cannot differentiate between a routine retrieval and an emergency departure, cannot provide reassurance or guidance to users unfamiliar with the process, and cannot proactively offer adaptive services when users encounter inconvenience. All interactions are based on commands and responses, lacking consideration for emotional dimensions, rendering the parking garage merely a functional space rather than a service-oriented intelligent environment, thus exhibiting insufficient intelligence. Summary of the Invention
[0004] The purpose of this invention is to provide a self-service car retrieval system for a vertically circulating intelligent parking garage, in order to solve the problems mentioned in the background art, such as indiscriminate scheduling, difficulty in recognizing user status, lack of emotional interaction, and insufficient intelligent service in existing vertically circulating intelligent parking garage self-service car retrieval systems.
[0005] To achieve the above objectives, the present invention provides the following technical solution: a vertical circulation three-dimensional intelligent parking garage self-service vehicle retrieval system, including a parking garage upper management platform, and a perception acquisition layer, a feature processing layer, a state determination layer, a mode triggering layer, a scheduling execution layer and a fair compensation layer connected to the parking garage upper management platform;
[0006] The perception and acquisition layer is used to acquire multi-dimensional status data of the user;
[0007] The feature processing layer parses the state data and extracts core features;
[0008] The state determination layer performs multi-feature fusion based on the core features to determine the user's state.
[0009] The mode triggering layer triggers the corresponding scheduling mode based on the user's status;
[0010] The scheduling execution layer executes the corresponding scheduling logic to complete vehicle scheduling;
[0011] The fair compensation layer is used to compensate users affected during the scheduling process.
[0012] As a preferred technical solution of the present invention, the perception acquisition layer includes a facial expression acquisition unit, a gait feature acquisition unit, a voice feature acquisition unit, and an operation behavior acquisition unit;
[0013] The facial expression acquisition unit uses a high-definition camera to capture micro-expression images of the user's face; the gait feature acquisition unit uses millimeter-wave radar to collect the user's gait parameters; the voice feature acquisition unit uses a microphone to collect the user's voice acoustic features; and the operation behavior acquisition unit obtains user operation behavior data through an operation terminal.
[0014] As a preferred technical solution of the present invention, the feature processing layer includes a facial micro-expression analysis module, a gait parameter extraction module, a speech acoustic analysis module, and an operation behavior analysis module;
[0015] The facial micro-expression analysis module extracts feature points and performs emotion correlation analysis on the acquired facial images. The gait parameter extraction module extracts the user's step frequency, stride length, and dwell time. The speech acoustic analysis module extracts the user's speech tone, speech rate, and intensity features. The operation behavior analysis module counts the user's operation hesitation time and operation frequency.
[0016] As a preferred technical solution of the present invention, the state determination layer includes a multi-feature fusion determination engine. The multi-feature fusion determination engine performs correlation analysis on facial emotion features, gait parameters, speech acoustic features and operation behavior features through a weighted fusion algorithm, and excludes spoofing behavior through abnormal feature screening, and outputs user state labels and state confidence. The user state labels include three categories: hurried, troubled and calm.
[0017] As a preferred technical solution of the present invention, the mode triggering layer includes a priority mode triggering module, a guide mode triggering module, and a basic scheduling mode triggering module; when the state determination layer determines that the user's state is urgent, the priority mode triggering module triggers the priority scheduling mode; when the user's state is determined to be troubled, the guide mode triggering module triggers the guide scheduling mode; when the user's state is determined to be calm, the basic scheduling mode triggering module triggers the regular timing scheduling mode.
[0018] As a preferred technical solution of the present invention, the scheduling execution layer includes an intelligent scheduling algorithm engine, a garage equipment control module, and a user interaction terminal control module; in the priority scheduling mode, the intelligent scheduling algorithm engine dynamically replans the vehicle retrieval path to reduce waiting time, the garage equipment control module opens the exit channel in advance and locks the safety protection mechanism, and the user interaction terminal control module informs the user of the priority processing status through screen display and voice.
[0019] As a preferred technical solution of the present invention, in the guided scheduling mode, the user interaction terminal control module switches to low-speed voice broadcast and step-by-step visual prompts, and continuously guides the user to complete the vehicle retrieval operation through the operation behavior collection unit to monitor the user's operation status in real time.
[0020] As a preferred technical solution of the present invention, the fair compensation layer includes an affected user identification module, a delay information push module, and a lightweight priority compensation queue management module; the affected user identification module identifies the first subsequent user affected during the scheduling process, the delay information push module pushes the expected delay duration and reason explanation to the affected user, and the lightweight priority compensation queue management module includes the next vehicle retrieval request of the affected user in the compensation queue and assigns it a weighted priority.
[0021] Compared with the prior art, the beneficial effects of the present invention are:
[0022] Through the design of this invention, the system can capture subtle changes in a user's facial expressions, gait rhythm, or tone of voice in a non-contact manner, thereby determining whether the user's emotional state is "urgent," "calm," or "disturbed." When the system detects that the user is in a hurry, it seamlessly activates "priority mode," re-planning the scheduling algorithm to reduce unnecessary waiting time and opening the exit channel in advance, while providing clear voice and interface feedback to inform the user that "priority processing is underway." For users who show confusion or pause, the system can switch to "guided mode," using a more patient speaking speed and more detailed visual cues to assist in the operation. Simultaneously, the system... By comprehensively analyzing multi-dimensional parameters that are difficult to fake, such as user gait frequency, operational interaction hesitation time, and objective acoustic characteristics of voice requests, the system can reliably determine whether a user is in a "normal," "urgent," or "confused" state. This effectively prevents the behavior of faking priority through simple facial expressions or actions. At the same time, the system will automatically send the estimated delay time and explanation to the next affected user and include that user's next request in the lightweight priority compensation queue. This not only improves the humanization and efficiency of service response in real emergency situations, but also ensures the overall scheduling fairness through technical means, eliminates the vulnerability of malicious queue jumping, and reduces user disputes caused by priority conflicts. Attached Figure Description
[0023] Figure 1 This is a system diagram of the present invention. Detailed Implementation
[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. 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.
[0025] Please see Figure 1 The present invention provides a technical solution: a vertical circulation three-dimensional intelligent parking garage self-service vehicle retrieval system, including a parking garage upper management platform, and a perception acquisition layer, feature processing layer, state determination layer, mode triggering layer, scheduling execution layer and fair compensation layer connected to the parking garage upper management platform;
[0026] The perception and acquisition layer is used to acquire multi-dimensional state data of users;
[0027] The feature processing layer parses the state data and extracts core features;
[0028] The state determination layer performs multi-feature fusion based on core features to determine the user's state;
[0029] The mode triggering layer triggers the corresponding scheduling mode based on the user's state;
[0030] The scheduling execution layer executes the corresponding scheduling logic to complete vehicle scheduling;
[0031] The fair compensation layer is used to compensate users affected during the scheduling process.
[0032] Furthermore, in actual operation, the system enables efficient collaboration between its various layers through standard data interfaces and communication protocols. The parking garage management platform, as the core coordinating unit, is responsible for uniformly configuring system parameters, recording operational logs, handling abnormal alarms, and providing administrators with a visual monitoring interface.
[0033] In this embodiment, the perception acquisition layer includes a facial expression acquisition unit, a gait feature acquisition unit, a voice feature acquisition unit, and an operation behavior acquisition unit;
[0034] The facial expression acquisition unit uses a high-definition camera to capture micro-expression images of the user's face; the gait feature acquisition unit uses millimeter-wave radar to collect the user's gait parameters; the voice feature acquisition unit uses a microphone to collect the user's voice acoustic features; and the operation behavior acquisition unit obtains user operation behavior data through the operation terminal.
[0035] Specifically, the perception and acquisition layer is automatically activated when the user approaches the garage operation area, and each acquisition unit starts synchronously. The time-series consistency of multimodal data is ensured through a timestamp alignment mechanism, providing a reliable data foundation for subsequent feature fusion.
[0036] In this embodiment, the feature processing layer includes a facial micro-expression analysis module, a gait parameter extraction module, a speech acoustic analysis module, and an operation behavior analysis module;
[0037] The facial micro-expression analysis module extracts feature points and performs emotion correlation analysis on the collected facial images; the gait parameter extraction module extracts the user's step frequency, stride length and dwell time; the speech acoustic analysis module extracts the user's speech tone, speech rate and intensity features; and the operation behavior analysis module counts the user's operation hesitation time and operation frequency.
[0038] In this embodiment, preferably, the facial micro-expression analysis module uses a convolutional neural network to achieve real-time feature point detection and expression classification; the gait parameter extraction module uses a radar signal processing algorithm to extract spatiotemporal gait features; the speech acoustic analysis module extracts acoustic feature vectors through Mel-frequency cepstral coefficients; and the operation behavior analysis module combines interface event flow to model behavior patterns.
[0039] In this embodiment, the state determination layer includes a multi-feature fusion determination engine. The multi-feature fusion determination engine uses a weighted fusion algorithm to perform correlation analysis on facial emotion features, gait parameters, speech acoustic features and operation behavior features, and eliminates spoofing behavior through abnormal feature screening, and outputs user state labels and state confidence. The user state labels include three categories: hurried, distressed and calm.
[0040] Specifically, the multi-feature fusion judgment engine adopts an adaptive weighting mechanism, which dynamically adjusts the weights based on the signal-to-noise ratio and real-time reliability of each feature. At the same time, a behavior consistency verification module is introduced. If a certain feature deviates significantly from the trend of other features, an anomaly screening process is triggered, thereby improving the robustness and anti-deception capability of state judgment.
[0041] In this embodiment, the mode triggering layer includes a priority mode triggering module, a guide mode triggering module, and a basic scheduling mode triggering module. When the state determination layer determines that the user's state is urgent, the priority mode triggering module triggers the priority scheduling mode. When the user's state is determined to be distressed, the guide mode triggering module triggers the guide scheduling mode. When the user's state is determined to be calm, the basic scheduling mode triggering module triggers the regular timing scheduling mode.
[0042] Furthermore, a confidence threshold judgment logic is set between the mode triggering layer and the state determination layer: the corresponding mode is triggered only when the state confidence is higher than the set threshold; otherwise, the system will run according to the basic scheduling mode and prompt the user to reconfirm the operation intention through the interactive terminal.
[0043] In this embodiment, the scheduling execution layer includes an intelligent scheduling algorithm engine, a garage equipment control module, and a user interaction terminal control module. In the priority scheduling mode, the intelligent scheduling algorithm engine dynamically replans the vehicle retrieval path to reduce waiting time, the garage equipment control module opens the exit channel in advance and locks the safety protection mechanism, and the user interaction terminal control module informs the user of the priority processing status through screen display and voice. Based on the real-time garage status (such as the position of the vehicle platform, motor load, and safety sensor status) and user status information, the intelligent scheduling algorithm engine uses a heuristic search algorithm to quickly generate a locally optimized scheduling sequence, and continuously monitors the system safety boundary during execution to ensure that priority scheduling does not affect the overall structural stability and personnel safety.
[0044] In this embodiment, under the guided scheduling mode, the user interaction terminal control module switches to low-speed voice broadcast and step-by-step visual prompts, and monitors the user's operation status in real time through the operation behavior collection unit, continuously guiding the user to complete the vehicle retrieval operation. In the guided scheduling mode, the system decomposes the operation process into multiple simple steps. After each step is completed, the user must confirm or the system must detect the correct operation before proceeding to the next step. If the user does not operate for a long time or makes multiple operational errors, the system will automatically transfer to the manual assistance channel.
[0045] In this embodiment, the fair compensation layer includes an affected user identification module, a delay information push module, and a lightweight priority compensation queue management module. The affected user identification module identifies the first subsequent user affected during the scheduling process. The delay information push module pushes the estimated delay time and reason to the affected user. The lightweight priority compensation queue management module includes the next vehicle retrieval request of the affected user in the compensation queue and assigns it a weighted priority. The lightweight priority compensation queue management module adopts a dynamic priority adjustment strategy, in which the priority weight of the compensated user increases with the number of times it waits and is automatically reset after it completes a priority vehicle retrieval. This reflects the principle of fair compensation and prevents the compensation mechanism from being occupied for a long time.
[0046] Although embodiments of the invention have been shown and described (see the detailed description above), it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.
Claims
1. A vertical circulation intelligent parking garage self-service vehicle retrieval system, characterized in that: It includes a parking garage management platform, and a perception acquisition layer, a feature processing layer, a state determination layer, a mode triggering layer, a scheduling execution layer, and a fair compensation layer connected to the parking garage management platform; The perception and acquisition layer is used to acquire multi-dimensional status data of the user; The feature processing layer parses the state data and extracts core features; The state determination layer performs multi-feature fusion based on the core features to determine the user's state. The mode triggering layer triggers the corresponding scheduling mode based on the user's status; The scheduling execution layer executes the corresponding scheduling logic to complete vehicle scheduling; The fair compensation layer is used to compensate users affected during the scheduling process.
2. The self-service car retrieval system for a vertical circulation intelligent parking garage according to claim 1, characterized in that: The perception and acquisition layer includes a facial expression acquisition unit, a gait feature acquisition unit, a voice feature acquisition unit, and an operation behavior acquisition unit; The facial expression acquisition unit uses a high-definition camera to capture micro-expression images of the user's face; the gait feature acquisition unit uses millimeter-wave radar to collect the user's gait parameters; the voice feature acquisition unit uses a microphone to collect the user's voice acoustic features; and the operation behavior acquisition unit obtains user operation behavior data through an operation terminal.
3. The self-service car retrieval system for a vertically circulating intelligent parking garage according to claim 1, characterized in that: The feature processing layer includes a facial micro-expression analysis module, a gait parameter extraction module, a speech acoustic analysis module, and an operation behavior analysis module; The facial micro-expression analysis module extracts feature points and performs emotion correlation analysis on the acquired facial images. The gait parameter extraction module extracts the user's step frequency, stride length, and dwell time. The speech acoustic analysis module extracts the user's speech tone, speech rate, and intensity features. The operation behavior analysis module counts the user's operation hesitation time and operation frequency.
4. The self-service car retrieval system for a vertical circulation intelligent parking garage according to claim 1, characterized in that: The state determination layer includes a multi-feature fusion determination engine. The multi-feature fusion determination engine uses a weighted fusion algorithm to perform correlation analysis on facial emotion features, gait parameters, speech acoustic features and operation behavior features, and excludes spoofing behavior through abnormal feature screening, and outputs user state labels and state confidence scores. The user state labels include three categories: hurried, distressed and calm.
5. The self-service car retrieval system for a vertical circulation intelligent parking garage according to claim 1, characterized in that: The mode triggering layer includes a priority mode triggering module, a guide mode triggering module, and a basic scheduling mode triggering module. When the state determination layer determines that the user's state is urgent, the priority mode triggering module triggers the priority scheduling mode. When the user's state is determined to be distressed, the guide mode triggering module triggers the guide scheduling mode. When the user's state is determined to be calm, the basic scheduling mode triggering module triggers the regular timing scheduling mode.
6. The self-service car retrieval system for a vertically circulating intelligent parking garage according to claim 5, characterized in that: The scheduling execution layer includes an intelligent scheduling algorithm engine, a garage equipment control module, and a user interaction terminal control module. In the priority scheduling mode, the intelligent scheduling algorithm engine dynamically replans the vehicle retrieval route to reduce waiting time, the garage equipment control module opens the exit channel in advance and locks the security protection mechanism, and the user interaction terminal control module informs the user of the priority processing status through screen display and voice.
7. The self-service car retrieval system for a vertically circulating intelligent parking garage according to claim 5, characterized in that: In the guided scheduling mode, the user interaction terminal control module switches to low-speed voice broadcasting and step-by-step visual prompts, and continuously guides the user to complete the vehicle retrieval operation through the operation behavior collection unit to monitor the user's operation status in real time.
8. The self-service car retrieval system for a vertical circulation intelligent parking garage according to claim 1, characterized in that: The fair compensation layer includes an affected user identification module, a delay information push module, and a lightweight priority compensation queue management module. The affected user identification module identifies the first subsequent user affected during the scheduling process. The delay information push module pushes the estimated delay duration and reason to the affected user. The lightweight priority compensation queue management module includes the affected user's next vehicle retrieval request in the compensation queue and assigns it a weighted priority.