Intelligent control system of modular office furniture

Abstract

The invention relates to the technical field of intelligent control, in particular to an intelligent control system of modular office furniture, which comprises a sitting posture analysis module, a height instruction generation module, an in-situ feature extraction module, a light instruction generation module and an adjustment execution module. According to the invention, by synchronously analyzing the continuous change data of the desktop height and the migration trend of the seat pressure distribution, the sitting posture state change of the user caused by work task switching can be accurately identified, and then the desktop is actively adjusted to the personalized preset height conforming to the current posture; dynamic human factor engineering adaptation without manual intervention is achieved, meanwhile, through comprehensive judgment of the average pressure, the pressure instantaneous fluctuation peak value and the static duration, a multi-dimensional in-place state recognition basis is formed, the basis can effectively distinguish real off-seat and transient posture fine adjustment of a user, and the user experience is improved. Therefore, frequent light starting and stopping caused by erroneous judgment are avoided, the use experience is improved, and energy is saved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent control, and in particular to an intelligent control system for modular office furniture. BACKGROUND

[0002] The technical field of intelligent control mainly involves the automatic control of various devices and systems, including but not limited to smart home, automated production lines, intelligent transportation, robot control, etc. Through sensors, controllers, and actuators, the system is intelligently adjusted and managed.

[0003] Among them, the intelligent control system for office furniture refers to the automatic adjustment of various functions of office furniture such as seat height and desk angle through the integration of sensors and electric drive devices. It usually relies on sensors to monitor user status or environmental changes and adjusts the settings of the furniture in real time to optimize the user's experience.

[0004] The intelligent control system of the prior art usually relies on single or non-associated sensor data for response, which has limitations in understanding the user's true intention. For example, when the user is only temporarily leaning back to relax or leaning forward to focus more on the screen, the system cannot actively match a more appropriate desk height because it cannot correlate the analysis of the sitting posture and the state of the desktop device, resulting in the user having to frequently manually adjust to maintain comfort, which disrupts the continuity of work and weakens the convenience of intelligentization. In addition, its in-place judgment logic is relatively simple and relies mainly on the presence or absence of basic pressure or movement detection. In actual office scenarios, the user may be incorrectly determined to be out of place when they briefly stand up to get something or significantly adjust their posture on the seat, causing the desktop light to unnecessarily turn off. When the user leaves after placing heavy objects on the chair surface, the system may also incorrectly maintain the light on, causing unnecessary energy waste. Its adjustment execution process is often a direct linear response, and once a deviation is detected, it adjusts in a fixed mode, which can appear rigid and interfere with the user. SUMMARY

[0005] The purpose of the present application is to solve the shortcomings in the prior art and to provide an intelligent control system for modular office furniture.

[0006] In order to achieve the above-mentioned purpose, the present application adopts the following technical scheme: an intelligent control system for modular office furniture comprises: A sitting posture analysis module acquires the desk height and pressure distribution of the office chair seat cushion in the current period through a sensor array, judges the user's sitting posture state according to the change trend of the desk height and pressure distribution at adjacent time points in the current period, and marks the sitting posture change type. The height instruction generation module compares the sitting posture change type with a preset fuzzy rule base, judges the lifting amplitude of the current office table, and generates an office table lifting instruction according to the lifting amplitude; The in-situ feature extraction module extracts a seat cushion average pressure value, a maximum instantaneous seat cushion pressure fluctuation amplitude, and a static duration in which a pressure change value is continuously maintained within a set time threshold range from the pressure distribution of the office chair seat cushion in the current period, and constructs a seat cushion state feature vector. The light instruction generation module judges the in-situ or off-situ state of the user according to the seat cushion state feature vector, and generates a desktop light brightness adjustment instruction according to the judgment result. The adjustment execution module adjusts the height of the office table desktop and the light brightness of the office table desktop according to the office table lifting instruction and the desktop light brightness adjustment instruction, and obtains an office furniture control result.

[0007] As a further scheme of the present application, the sitting posture change type includes a body leaning forward state marker, a body leaning backward state marker, and a sitting posture deviation state marker, the office table lifting instruction specifically includes an upward instruction, a downward instruction, and a keep original height instruction, the seat cushion state feature vector includes an in-situ stable feature component, an off-situ change feature component, and a transition state feature component, the desktop light brightness adjustment instruction includes an enhanced brightness instruction, a reduced brightness instruction, and a keep brightness instruction, and the office furniture control result includes an adjusted desktop height and an adjusted light brightness.

[0008] As a further scheme of the present application, the sitting posture analysis module includes: The desktop height acquisition submodule acquires the output value of the height sensor arranged at the desktop position of the office table, collects the continuous change data of the desktop height of the office table at a set time interval in the current period, and generates a desktop height change sequence. The seat cushion pressure acquisition submodule acquires the output value of the pressure sensor arranged in the seat cushion area of the office chair, collects the pressure distribution of the seat cushion of the office chair in the same time period as the desktop height, extracts the force area position and change at each time point, and generates a seat cushion pressure distribution change sequence. The sitting posture change judgment submodule calls the desktop height change sequence and the seat cushion pressure distribution change sequence, judges whether there is a trend combination of synchronous association according to the displacement trend of the desktop height and the front and back change direction of the seat cushion pressure area between adjacent time points, and then marks the sitting posture state change and generates a sitting posture change type.

[0009] As a further scheme of the present application, the trend combination is the upward or downward trend of the desktop height of the office table and the forward movement, backward movement, or pressure lowering change of the force concentration position of the seat cushion of the office chair between adjacent time points.

[0010] As a further scheme of the present application, the height instruction generation module comprises: a sitting posture matching sub-module, which generates a desk height deviation value according to a comparison between a target reference height of a desk top corresponding to the current sitting posture change type and a measured value of the desk height sensor in the current period; a height deviation calculation sub-module, which calculates a numerical range of the desk height deviation value, controls the size of the membership degree in the fuzzy rule base within the range of the control amplitude, selects the control amplitude with the largest membership degree as a target adjustment range, and generates a desk lifting action amplitude value; a lifting instruction generation sub-module, which calls the desk lifting action amplitude value, combines a corresponding action direction and the target change range, and generates a desk lifting instruction.

[0011] As a further scheme of the present application, the fuzzy rule base is set by statistically recording the desk top height deviation data of the user in a continuous sitting posture change process and observing and recording the response degree of the user to the desk height in an actual adjustment process.

[0012] As a further scheme of the present application, the in-situ feature extraction module comprises: an average pressure calculation sub-module, which obtains pressure distribution values output by pressure sensors arranged in a seat cushion area of an office chair at each time point in the current period, and calculates seat cushion average pressure values of all the sensors at each time point; a fluctuation amplitude identification sub-module, which calls a pressure data sequence of the office chair seat cushion pressure sensor in the current period, calculates pressure change difference values between adjacent time points in the sequence, identifies the maximum value of the difference values and records the maximum value as a maximum instantaneous seat cushion pressure fluctuation amplitude; a static state judgment sub-module, which calls the pressure data sequence of the office chair seat cushion pressure sensor in the current period, judges the length of a continuous time section within a set pressure fluctuation threshold range, selects a section with the longest time span as a static duration, and integrates the seat cushion average pressure value and the maximum instantaneous seat cushion pressure fluctuation amplitude to construct a seat cushion state feature vector.

[0013] As a further scheme of the present application, the light instruction generation module comprises: a feature calling sub-module, which calls the seat cushion state feature vector as an input basis for in-situ state recognition, and generates a light control judgment input data set; an in-situ state judgment sub-module, which judges the in-situ state or off-situ state of the user according to the light control judgment input data set, and generates a current seat state judgment result of the user; a brightness instruction generation sub-module, which marks a corresponding adjustment amplitude range according to the current seat state judgment result of the user, and generates a desk light brightness adjustment instruction.

[0014] As a further scheme of the present application, the judging the in-position state or the off-position state of the user is comparing the average seat cushion pressure value with an in-position pressure threshold value, comparing the maximum instantaneous seat cushion pressure fluctuation amplitude with a fluctuation threshold value, comparing the stationary duration time with a time threshold value, if all three conditions are met, the in-position state is determined, if any one condition is not met, the off-position state is determined.

[0015] As a further scheme of the present application, the adjusting execution module comprises: The lifting control execution submodule outputs a control signal to the office table lifting driving structure according to the office table lifting instruction, and generates a desktop height adjustment result; The light control execution submodule outputs a control signal to the office table light control structure according to the desktop light brightness adjustment instruction, and generates a desktop light adjustment result; The data updating submodule integrates the desktop height adjustment result and the desktop light adjustment result, updates the adjusted desktop height and light brightness parameters, and obtains an office furniture control result.

[0016] Compared with the prior art, the present application has the following advantages and positive effects: In the present application, by synchronously analyzing the continuous change data of the desktop height and the migration trend of the seat pressure distribution, the change of the sitting posture state of the user caused by the switching of the work task can be accurately recognized, and then the desktop is actively adjusted to the personalized preset height conforming to the current posture, realizing the dynamic human factors engineering adaptation without manual intervention. At the same time, through the comprehensive judgment of the average pressure, the instantaneous fluctuation peak value of the pressure and the stationary duration time, a multi-dimensional in-position state recognition basis is formed, which can effectively distinguish the real off-seat of the user and the temporary posture fine adjustment, avoid the frequent start and stop of the light caused by misjudgment, improve the use experience and save energy. Finally, when adjusting the desktop height, through the multi-interval fuzzy evaluation of the height deviation, a smooth and nonlinear lifting amplitude is generated, so that the adjustment process is more natural and soft, and the mechanical abrupt lifting action is avoided. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 The system flowchart of the present application is shown in the figure; Figure 2 The flowchart of the sitting posture analysis module of the present application is shown in the figure; Figure 3 The flowchart of the height instruction generation module of the present application is shown in the figure; Figure 4 The flowchart of the in-position feature extraction module of the present application is shown in the figure; Figure 5 The flowchart of the light instruction generation module of the present application is shown in the figure; Figure 6This is a flowchart of the adjustment and execution module of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0019] Please see Figure 1 An intelligent control system for modular office furniture includes: The posture analysis module uses a sensor array to collect data on the height of the office desk and the pressure distribution of the office chair seat within the current period. Based on the changing trends of the height of the desk and the pressure distribution at adjacent time points within the current period, it determines the user's posture and marks the type of posture change. The height command generation module compares the type of sitting posture change with a preset fuzzy rule library to determine the current height adjustment range of the desk, and generates desk height adjustment commands based on the height adjustment range. The in-situ feature extraction module extracts the average pressure value of the office chair seat cushion, the maximum instantaneous pressure fluctuation amplitude of the seat cushion, and the static duration during which the pressure change value remains within a set time threshold from the pressure distribution of the office chair seat cushion in the current period, and constructs a seat cushion state feature vector. The lighting instruction generation module determines the user's current position (in or out) based on the seat cushion's state feature vector, and generates a desktop lighting brightness adjustment instruction based on the determination result. The adjustment and execution module adjusts the height of the office desk and the brightness of the desk lights according to the office desk height adjustment command and the desk light brightness adjustment command, so as to obtain the office furniture control result; The types of sitting posture changes include forward leaning state marking, backward leaning state marking, and sitting posture deviation state marking. The specific commands for raising and lowering the desk are upward command, downward command, and maintain original height command. The seat cushion state feature vector includes in-position stable feature component, displacement change feature component, and transition state feature component. The desktop light brightness adjustment commands include brightness increase command, brightness decrease command, and maintain brightness command. The office furniture control results include the adjusted desktop height and the adjusted light brightness.

[0020] Please see Figure 2 The sitting posture analysis module includes: The desktop height acquisition submodule acquires the output value of the height sensor installed on the desktop of the office desk, collects continuous change data of the desktop height at set time intervals within the current period, and generates a desktop height change sequence. The output values are obtained by ToF laser ranging sensors arranged in the four corners under the desktop of the office table. The main controller sends data collection instructions to each laser ranging sensor at a fixed time interval in a preset analysis period through I2C bus in a polling manner. After the sensor completes the measurement, the vertical height data of the desktop from the ground is returned to the main controller. The main controller marks each set of four collected height values with an accurate time stamp, compiles the four height values with the same time stamp into a data frame containing the coordinates of four points in space, and arranges all the data frames in time sequence by continuously performing the above collection and compilation actions in the analysis period, thereby forming an original data sequence that can accurately describe the continuous change of the height and tilt posture of the desktop of the office table in the time dimension. The sequence not only records the overall lifting of the desktop, but also reflects the slight tilt change caused by uneven force. Finally, the height change sequence of the desktop of the office table is generated.

[0021] The seat cushion pressure acquisition sub-module obtains the output values of the pressure sensors arranged in the seat cushion area of the office chair, collects the pressure distribution of the seat cushion of the office chair in the same time period as the desktop height, extracts the force area position and change at each time point, and generates a seat cushion pressure distribution change sequence. The output values of the array composed of 256 MEMS piezoresistive sensors arranged under the seat cushion of the office chair are obtained. In the same analysis period as the desktop height acquisition, the seat cushion controller synchronously scans the rows and columns of the sensor matrix to construct a complete pressure distribution map. For each collected pressure distribution map, the data processing unit first compares the reading of each sensor in the map with a preset contact threshold value. The contact threshold value is set based on the statistical analysis of the background noise of the sensor array for several hours in the unloaded state of the office chair. The pressure value that can filter out most (for example, 99.7%) noise signals is selected. The sensor with a reading lower than the threshold value is considered as not effectively contacted, and its value will be ignored or set to zero. Subsequently, the data processing unit obtains the force center position at the current time by performing pressure value weighted average calculation on the coordinates of all effective pressure points. At the same time, the region composed of sensors with pressure readings exceeding a high pressure threshold value is identified. The high pressure threshold value is obtained by statistically analyzing the pressure data of the core pressure area of the ischial tuberosities of hundreds of users with different body types in a standard sitting posture, and the 95th percentile of the pressure distribution is selected as the set value to accurately define the skeletal support points. Finally, the force center position coordinates at each time point, the area, shape and migration track of the high pressure region in the period are compiled into a sequence to generate the seat cushion pressure distribution change sequence of the office chair.

[0022] The sitting posture change determination sub-module calls the office table desktop height change sequence and the office chair seat cushion pressure distribution change sequence, and determines whether there is a synchronous associated trend combination according to the displacement trend of the desktop height and the front and back change direction of the seat cushion pressure area between adjacent time points. If yes, it is marked as a sitting posture state change, a sitting posture change type is generated, and the trend combination is the rising or falling trend of the office table desktop height between adjacent time points and the front movement, rear movement or stress concentration position depression change of the office chair seat cushion stress area. The processing unit first performs a first-order difference operation on the height value representing the geometric center of the desktop in the desktop height sequence to obtain the change rate of the desktop height with time, so as to determine the rising, falling or stable trend of the desktop. At the same time, the processing unit performs the same first-order difference operation on the front and back direction (usually Y-axis) coordinates of the stress center position in the seat cushion pressure sequence to determine the forward, backward or stable trend of the user's body center of gravity. Then, the change trend of the desktop height and the change direction of the seat cushion pressure center of gravity are combined into a feature pair, which is matched and queried in a pre-set trend combination rule library stored in a hash table structure. The rule library is established by inviting a large number of testers to wear inertial measurement units (IMU) to capture their body postures, synchronously recording the desktop and seat cushion data when performing various typical office actions (such as typing, reading, leaning back, standing up), and finding out the high-confidence association mode between the body posture and the data change through the association rule mining algorithm in machine learning. If the current detected feature pair is queried in the hash table to find a matching entry, it is determined that a sitting posture state change has occurred, and the corresponding sitting posture change type is generated according to the category defined by the entry.

[0023] Please refer to Figure 3 , the height instruction generation module includes: The sitting posture matching sub-module generates a desktop height deviation value by comparing the office table desktop target reference height corresponding to the current sitting posture change type with the actual measurement of the office table height sensor in the current period. According to the current generated sitting posture change type, the type is taken as a query key to search in a user sitting posture preference database stored in a local non-volatile flash memory. The database is a relational data table, and the user actively fills in the table by manually adjusting and confirming the desktop height that feels most ergonomic in different working states (such as “intensive reading” and “regular typing”). The query operation returns the target reference height of the office table desktop uniquely corresponding to the current sitting posture change type. After obtaining the target reference height, the processing unit performs an arithmetic average calculation on all valid readings of the office table height sensor collected in the current analysis period to obtain a measured height value that can filter out transient measurement errors and represents the stable height of the current desktop. Finally, the processing unit performs a subtraction operation on the measured height value and the target reference height value retrieved from the database to generate a desktop height deviation value.

[0024] The height deviation calculation submodule calculates the numerical range of the desktop height deviation value, controls the size of the membership degree within the range of the control amplitude in the fuzzy rule library, selects the control amplitude with the largest membership degree as the target adjustment range, generates the office table lifting action amplitude value, and sets the fuzzy rule library by statistically analyzing the office table desktop height deviation data corresponding to the user in the continuous sitting posture change process and combining the observation records of the user's response to the office table height in the actual adjustment process. The fuzzy set used to describe the degree of desktop height deviation is specifically divided into five categories, each of which corresponds to a specific deviation interval: NegativeBig (NB): indicates that the current height of the desktop is much lower than the user's target height. This set mainly covers the deviation interval (-∞, -4.0] cm.

[0025] NegativeSmall (NS): indicates that the current height of the desktop is slightly lower than the user's target height. This set mainly covers the deviation interval (-4.0, -1.0] cm.

[0026] Zero (ZE): indicates that the current height of the desktop is basically at or very close to the user's target height, and does not need to be adjusted significantly. This set mainly covers the deviation interval (-1.0, 1.0] cm.

[0027] PositiveSmall (PS): indicates that the current height of the desktop is slightly higher than the user's target height. This set mainly covers the deviation interval (1.0, 4.0] cm.

[0028] PositiveBig (PB): indicates that the current height of the desktop is much higher than the user's target height. This set mainly covers the deviation interval (4.0, +∞) cm.

[0029] The boundaries of the intervals are flexible in fuzzy logic, and there are overlapping transition zones. In the actual judgment process, a specific deviation value (for example, +3.5 cm) will be compared with all fuzzy sets at the same time to calculate the degree of compliance with each set description. The system will determine its membership degree according to the proximity of the deviation value to the typical center value of each set, so a deviation value may mainly belong to "small positive deviation", but also slightly belong to "large positive deviation".

[0030] First, the membership of the desktop height deviation value to different fuzzy sets is calculated by the Gaussian membership function, which is expressed as In the function: is the membership, which is a dimensionless value between 0 and 1, representing the degree of compliance of the desktop height deviation value with the fuzzy set described; is the difference between the actual height of the desktop and the target reference height, which is a physical quantity with positive and negative signs, and the unit is centimeter; is the base of natural logarithm, which is a dimensionless constant; is the center value of the fuzzy set , representing the most typical deviation value of the fuzzy set, with the unit of centimeter. The center value is set by K-means clustering analysis on a large number of user historical manual adjustment data, taking the centroid (i.e. data center point) of each cluster as the set value; is the identification of the fuzzy set, used to distinguish different degrees of deviation; is the standard deviation of the fuzzy set , which determines the shape of the membership function curve and reflects the distribution width of the data points around the center value, with the unit of centimeter. The standard deviation is set according to the standard deviation of the distance of all data points in the corresponding cluster to the centroid. After determining the fuzzy set with the highest membership, the office table lifting action amplitude value is calculated by adjusting the amplitude calculation formula , In the formula: is the final generated desktop lifting adjustment amount, with the unit of centimeter, and the positive and negative signs represent the lifting direction; is the proportional control weight coefficient, which is a dimensionless parameter, determining the direct contribution of the original deviation value in the final adjustment amount; is the fuzzy control weight coefficient, which is a dimensionless parameter, determining the contribution of the deviation part after fuzzy logic fine-tuning in the final adjustment amount, and The numerical setting is based on user feedback and the optimization target of adjustment performance, and is derived through least squares regression analysis of historical adjustment data. If the analysis shows that users' adjustment behavior tends to quickly correct larger deviations, then... The weight of [the specific feature] will increase accordingly; conversely, if the user prefers fine-tuning, then [the weight of the feature] will decrease. The weight will increase.

[0031] Set the desktop height deviation value The value is +3.51 cm. Two related fuzzy sets are defined in the fuzzy rule base: "Positive Small Deviation (PS)" and "Positive Large Deviation (PB)". The center value of "Positive Small Deviation (PS)" is set based on historical data analysis. The value is 2.0 cm, and the standard deviation is 2.0 cm. The value is 1.5 cm; the center value of "positive large deviation (PB)" The value is 6.0 cm, and the standard deviation is 6.0 cm. It is 2.0 cm.

[0032] The calculation process for the first formula: First, calculate the deviation value respectively. For the two fuzzy sets "Positive Small Deviation (PS)" and "Positive Large Deviation (PB)", calculate the membership degree for "Positive Small Deviation (PS)". :Will , , Substitute into the formula: 2. Calculate the membership degree for "positive large deviation (PB)". :Will , , Substitute into the formula: Compare two membership values ​​because Therefore, the deviation value This better aligns with the description of "positive small deviation (PS)". Therefore, subsequent calculations will use the fuzzy set "positive small deviation (PS)" and related parameters.

[0033] The calculation process for the second formula: Next, calculate the final value of the desk height adjustment range. Based on historical data regression analysis, weighting coefficients are set. It is 0.6. The value is 0.4. The known value is... centimeter, , And the membership degree calculated in the previous step. and the corresponding center value Substitute centimeters into the formula: Calculation results about -2.47 centimeters, indicating that a lifting instruction should be generated to lower the desktop by 2.47 centimeters.

[0034] The lifting instruction generation submodule calls the office table lifting action amplitude value, combines the corresponding action direction and target variation range, and generates an office table lifting instruction. The processing unit first determines the lifting action direction by judging the sign of the value, and if the value is negative, the preset direction is downward, and if the value is positive, the preset direction is upward. Then, the processing unit extracts the absolute value of the value as the target displacement of the lifting action. The displacement represents the physical distance that the desktop needs to move. Finally, the determined action direction and target displacement are encoded into a CAN bus message according to the preset communication protocol. The arbitration field ID of the message is used to identify the source and priority of the instruction. The specific byte bits in the data field are used to represent the direction (e.g., 0x01 for downward and 0x02 for upward). The remaining byte bits are used to represent the displacement data after unit conversion (e.g., from centimeters to 0.1 millimeters), to ensure that the drive controller can accurately parse and generate the office table lifting instruction.

[0035] Please refer to Figure 4 The position feature extraction module includes: The average pressure calculation submodule obtains the pressure distribution value output by the pressure sensor arranged in the seat cushion area of the office chair at each time point in the current period, and calculates the seat cushion average pressure value of all sensors at each time point. For each sampling time point in the analysis period, the data processing unit adds up the effective pressure readings of the 256 sensors after noise filtering, and then divides the total pressure value by the number of effective contact sensors (not the total number of 256) to calculate the seat cushion average pressure value at a single time point that better reflects the real contact area pressure condition. After completing data processing for all sampling time points in the period, a time series consisting of multiple instantaneous average pressure values is obtained, and then an arithmetic average operation is performed on all values in the sequence to obtain a final seat cushion average pressure value that can smooth the pressure fluctuations caused by small actions in the period and represent the overall pressure level of the current period.

[0036] The fluctuation amplitude recognition submodule calls the pressure data sequence of the office chair seat cushion pressure sensor in the current period, calculates the pressure change difference between adjacent time points in the sequence, identifies the maximum value of the difference, and records it as the maximum instantaneous seat cushion pressure fluctuation amplitude. The data processing unit generates a new difference sequence composed of the absolute values of the instantaneous change rate of pressure by performing subtraction operation between the values of adjacent time points in the sequence one by one and taking the absolute value, and each value in the sequence represents the intensity of the change of the average pressure of the seat cushion in a very short time interval. Then, the processing unit applies a peak detection algorithm to traverse the difference sequence, finds and determines an element with the maximum value by comparing one by one, and records the maximum value as the maximum instantaneous seat cushion pressure fluctuation amplitude, which represents the most intense pressure change event in the period.

[0037] The stationary state judgment submodule calls the pressure data sequence of the office chair seat cushion pressure sensor in the current period, judges the length of the continuous time section within the set pressure fluctuation threshold range, selects the longest time span as the stationary duration, and integrates the average seat cushion pressure value and the maximum instantaneous seat cushion pressure fluctuation amplitude to construct a seat cushion state feature vector. The stationary state judgment submodule calls the pressure data sequence of the office chair seat cushion pressure sensor in the current period, judges the length of the continuous time section within the set pressure fluctuation threshold range, selects the longest time span as the stationary duration, and integrates the average seat cushion pressure value and the maximum instantaneous seat cushion pressure fluctuation amplitude to construct a seat cushion state feature vector.

[0038] Please refer to Figure 5 , the light instruction generation module comprises: The feature calling submodule calls the seat cushion state feature vector as the input basis for the in-place state recognition, and generates a light control judgment input data set. The calling seat cushion state feature vector, the processing unit will contain the seat cushion average pressure value, the maximum instantaneous seat cushion pressure fluctuation amplitude, and the longest stationary duration in the vector, according to the predefined order and data type, encapsulate into a structured data package, the data package is loaded into the cache of the processor or the specified RAM address space, as the direct data source for subsequent bit state judgment submodule logic operation, so as to generate light control judgment input data group.

[0039] The bit state judgment submodule, according to the light control judgment input data group, judges the user's in-place state or off-site state, generates the user's current seat state judgment result, and judges the user's in-place state or off-site state by comparing the seat cushion average pressure value with the in-place pressure threshold, comparing the maximum instantaneous seat cushion pressure fluctuation amplitude with the fluctuation threshold, and comparing the stationary duration with the time threshold. If all three conditions are met, it is determined to be in place, and if any condition is not met, it is determined to be off-site. According to the light control judgment input data group, three independent logical comparison operations are performed in parallel. First, the seat cushion average pressure value is compared with the in-place pressure threshold. The threshold is set based on the seating experiment on the test population covering different weight ranges from teenagers to adults, recording the lowest seat cushion average pressure after sitting completely, and multiplying a coefficient less than 1 (for example, 0.8) as a safety margin to ensure that it can cover users with lighter weight. Second, the maximum instantaneous seat cushion pressure fluctuation amplitude is compared with the fluctuation threshold. The fluctuation threshold is determined by analyzing the pressure drop data when the user performs a quick standing and leaving action, and selecting a critical change rate that can distinguish normal sitting posture adjustment from standing action with high confidence (for example, 99%). Third, the stationary duration is compared with the time threshold. The time threshold is set based on the behavior analysis of the video of the body posture adjustment or short stationary time usually accompanied by a large number of users before leaving, and the critical length that can distinguish accidental pause from explicit intention to leave is counted. If the results of the above three comparisons meet the preset logical conditions (i.e. the average pressure is greater than the threshold, the fluctuation amplitude is less than the threshold, and the stationary time is greater than the threshold), it is determined to be in place, and the user's current seat state judgment result is generated.

[0040] The brightness instruction generation submodule, according to the user's current seat state judgment result, marks the corresponding adjustment range, and generates the desktop light brightness adjustment instruction. According to the current seat state determination result, a lookup is performed in a brightness control rule set stored in firmware and implemented in a decision tree structure. The rule set branches the "in-seat state" to a dynamic brightness adjustment logic and branches the "out-of-seat state" to a fixed low-power consumption brightness value (e.g., 10%). When the determination result is "in-seat state", the processing unit further acquires a current ambient illuminance value measured by an ambient light sensor integrated with the office table and color calibrated. A non-linear mapping function (e.g., a logarithmic function) is used to calculate a target brightness percentage complementary to the ambient illuminance in a brightness adjustment range of 80% to 100% to form optimal visual comfort. Finally, the rated maximum lumen value of the office table lamp is multiplied by the percentage to obtain a target brightness value, which is converted into an 8-bit PWM duty cycle value in the range of 0 to 255 to generate a desktop lamp brightness adjustment instruction.

[0041] Please refer to Figure 6 The adjustment execution module includes: The lifting control execution submodule outputs a control signal to the office table lifting drive structure according to the office table lifting instruction to generate a desktop height adjustment result. According to the office table lifting instruction, the CAN bus transceiver sends an instruction message to the motor controller of the office table lifting drive structure. After receiving the message and completing the cyclic redundancy check (CRC) without error, the motor controller parses the action direction and target displacement amount contained therein, and then precisely drives the three-phase direct-current brushless motor in the lifting column to start smooth operation through a complex H-bridge circuit based on a magnetic field-oriented control algorithm. At the same time, the high-precision absolute value encoder integrated at the tail of the motor feeds back the precise angular position of the rotor to the motor controller in real time. The controller accurately converts the angular position information into the actual displacement of the desktop according to the preset deceleration ratio and the lead of the high-precision ball screw, and compares it with the target displacement amount in a closed loop. When the actual displacement is equal to the target displacement amount, the controller immediately stops driving the motor and activates the electromagnetic brake, thereby completing the non-jittering and high-precision adjustment of the desktop height to generate a desktop height adjustment result.

[0042] The light control execution submodule outputs a control signal to the office table light control structure according to the desktop lamp brightness adjustment instruction to generate a desktop lamp adjustment result. According to the desktop light brightness adjustment instruction, the 8-bit PWM duty cycle value is sent to the light control microprocessor through the I2C serial bus, and after the microprocessor receives the value, the internal high-resolution timer / counter module is configured to generate a frequency stable (for example, above 1kHz to avoid human eye flicker) and accurate duty cycle PWM signal, the PWM signal is sent to the dimming control pin of the multi-channel constant current LED drive chip, and the drive chip adjusts the driving current of each channel of the desktop LED light strip according to the duty cycle of the input signal, so that the actual luminous intensity and color temperature (if adjustable) of the desktop light are linearly adjusted to the target state required by the instruction, and the desktop light adjustment result is generated.

[0043] The data updating submodule integrates the desktop height adjustment result and the desktop light adjustment result, updates the adjusted desktop height and light brightness parameters, and obtains the office furniture control result. The desktop height adjustment result and the desktop light adjustment result are integrated, and the latest adjusted desktop actual height value and light brightness parameters are updated into a specific data structure for storing the real-time running state of the device in the RAM through the direct memory access (DMA) mode. Simultaneously, in order to realize the power-off memory function, an asynchronous task of synchronously writing the key parameters in the data structure to the on-board EEPROM or Flash memory is triggered, the task adopts a wear leveling algorithm to prolong the service life of the storage medium, and the process ensures that the basis state data for the next intelligent adjustment is accurate, real-time and persistent, and the office furniture control result is obtained.

[0044] The above is only a preferred embodiment of the present application, and does not limit the form of the present application, any skilled person in the art can modify or change the above disclosed technical content to equivalent embodiments applied to other fields, but any simple modification, equivalent change and modification made according to the technical essence of the present application to the above embodiments without departing from the technical solution content of the present application still belongs to the protection scope of the present application.

Claims

1. A smart control system for modular office furniture, characterized in that, The system comprises: The sitting posture analysis module collects the desktop height and the pressure distribution of the office chair seat cushion in the current period through the sensor array, judges the user's sitting posture state according to the change trend of the desktop height and the pressure distribution at adjacent time points in the current period, and marks the sitting posture change type; The height instruction generation module compares the sitting posture change type with the preset fuzzy rule base, judges the lifting amplitude of the current office table, and generates the office table lifting instruction according to the lifting amplitude; The in-position feature extraction module extracts the seat cushion average pressure value, the maximum instantaneous seat cushion pressure fluctuation amplitude, and the static duration that the pressure change value continuously remains in the set time threshold range from the pressure distribution of the office chair seat cushion in the current period, and constructs a seat cushion state feature vector; The light instruction generation module judges the user's current in-position or off-position state according to the seat cushion state feature vector, and generates a desktop light brightness adjustment instruction according to the judgment result; The adjustment execution module adjusts the office table desktop height and the light brightness of the office table desktop according to the office table lifting instruction and the desktop light brightness adjustment instruction, and obtains the office furniture control result.

2. The intelligent control system of modular office furniture according to claim 1, characterized in that, The sitting posture change type includes body leaning forward state marking, body leaning backward state marking, and sitting posture deviation state marking, the office table lifting instruction is specifically an upward instruction, a downward instruction, or a keep original height instruction, the seat cushion state feature vector includes in-position stable feature component, off-position change feature component, and transition state feature component, the desktop light brightness adjustment instruction includes enhanced brightness instruction, reduced brightness instruction, and keep brightness instruction, and the office furniture control result includes adjusted desktop height and adjusted light brightness.

3. The intelligent control system of modular office furniture according to claim 1, characterized in that, The sitting posture analysis module comprises: The desktop height acquisition submodule acquires the output value of the height sensor arranged at the desktop position of the office table, collects the continuous change data of the office table desktop height at a set time interval in the current period, and generates an office table desktop height change sequence; The seat cushion pressure acquisition submodule acquires the output value of the pressure sensor arranged in the seat cushion area of the office chair, collects the pressure distribution of the office chair seat cushion in the same time period as the desktop height, extracts the stress area position and change at each time point, and generates an office chair seat cushion pressure distribution change sequence; The sitting posture change judgment submodule calls the office table desktop height change sequence and the office chair seat cushion pressure distribution change sequence, judges whether there is a synchronous associated trend combination according to the displacement trend of the desktop height and the front and back change direction of the seat cushion pressure area between adjacent time points, and then marks the sitting posture state change and generates the sitting posture change type.

4. The intelligent control system of modular office furniture according to claim 3, characterized in that, The trend combination is the rising or falling trend of the office table desktop height and the forward movement, backward movement, or pressure lowering change of the stress area of the office chair seat cushion between adjacent time points.

5. The intelligent control system of modular office furniture according to claim 1, characterized in that, The height instruction generation module comprises: The sitting posture matching submodule compares the current desktop height sensor measurement with the office table desktop target reference height corresponding to the current sitting posture change type, and generates a desktop height deviation value; A height deviation calculation sub-module calculates a numerical range of the desktop height deviation value, controls the size of the membership degree within the range of the control amplitude in the fuzzy rule base, selects the control amplitude with the largest membership degree as the target adjustment range, and generates an office table lifting action amplitude value; A lifting instruction generation sub-module calls the office table lifting action amplitude value, combines the corresponding action direction and the target variation range, and generates an office table lifting instruction.

6. The intelligent control system of modular office furniture according to claim 5, characterized in that, The fuzzy rule base is set by statistically recording the corresponding office table desktop height deviation data of the user in the continuous sitting posture change process and observing and recording the user's response to the office table height in the actual adjustment process.

7. The intelligent control system of modular office furniture according to claim 1, characterized in that, The in-place feature extraction module includes: An average pressure calculation sub-module acquires the pressure distribution value output by the pressure sensor arranged in the seat cushion area of the office chair at each time point in the current period, calculates the seat cushion average pressure value of all sensors at each time point, and identifies the maximum value of the difference value between the pressure change at adjacent time points in the sequence and records it as the maximum instantaneous seat cushion pressure fluctuation amplitude. A static state judgment sub-module calls the pressure data sequence of the office chair seat cushion pressure sensor in the current period, judges the length of the continuous time section within the set pressure fluctuation threshold range, selects the longest one as the static duration, and integrates the seat cushion average pressure value and the maximum instantaneous seat cushion pressure fluctuation amplitude to construct a seat cushion state feature vector. The light instruction generation module includes:

8. The intelligent control system of modular office furniture of claim 1, wherein, A feature calling sub-module calls the seat cushion state feature vector as the input basis for in-place state recognition, generates a light control judgment input data set, and judges the in-place state or off-site state of the user according to the light control judgment input data set, generates a user current seat state judgment result, and generates a desktop light brightness adjustment instruction according to the user current seat state judgment result. The judgment of the in-place state or off-site state of the user is to compare the seat cushion average pressure value with the in-place pressure threshold, compare the maximum instantaneous seat cushion pressure fluctuation amplitude with the fluctuation threshold, and compare the static duration with the time threshold. If all three conditions are met, it is determined to be in place, and if any condition is not met, it is determined to be off-site. The adjustment execution module includes: A lifting control execution sub-module outputs a control signal to the office table lifting drive structure according to the office table lifting instruction, generates a desktop height adjustment result, and a light control execution sub-module outputs a control signal to the office table light control structure according to the desktop light brightness adjustment instruction, generates a desktop light adjustment result.

9. The intelligent control system of modular office furniture according to claim 8, characterized in that, A data updating sub-module integrates the desktop height adjustment result and the desktop light adjustment result, updates the adjusted desktop height and light brightness parameters, and obtains an office furniture control result.

10. The intelligent control system of modular office furniture of claim 1, wherein, ​ ​ ​ ​

Images