Flavoring double-arm robot system based on three-dimensional machine vision

Abstract

The invention relates to the technical field of aromatherapy modulation, and provides a three-dimensional machine vision-based flavoring double-arm robot system, which comprises a flavoring double-arm robot, and is characterized in that the flavoring double-arm robot is provided with a three-dimensional machine vision module, a flavoring algorithm module and a mechanical arm provided with a clamping piece; the fragrance blending algorithm module is used for acquiring the aromatherapy demand of a user, and generating a plurality of aromatherapy formulas for the user to select based on the characteristics of aromatherapy raw materials and the aromatherapy demand of the user; the three-dimensional machine vision module is used for identifying and positioning the positions of required aromatherapy raw materials and a flavoring container after a user selects at least one aromatherapy formula, and generating a flavoring instruction for driving the mechanical arm and the clamping piece; the flavoring instruction is used for driving the mechanical arm and the clamping piece to execute weighing, charging, liquid transferring and stirring operations according to the required aromatherapy raw materials, the position of the flavoring container, the aromatherapy formula set proportion and the flavoring sequence so as to obtain the prepared aromatherapy, bottling the prepared aromatherapy, packaging and outputting the packaged aromatherapy to a user.

Description

Technical Field

[0001] The invention relates to the technical field of aromatherapy modulation, in particular to a fragrance-adjusting dual-arm robot system based on three-dimensional machine vision. Background Art

[0002] As consumers pursue personalized and high-quality life, the perfume industry needs to continuously launch innovative products that meet market demand. The preparation of traditional aromatherapy usually relies on the experience and manual operation of perfumers. The perfume making process includes the selection, proportioning and mixing of raw materials. The perfume making process often requires multiple tests and adjustments to ensure the concentration and fragrance of the aromatherapy. Due to its high dependence and labor intensity, the efficiency is low. Perfumers will constantly evaluate the fragrance of the aromatherapy and make adjustments as needed. This process requires repeated experiments to achieve the desired effect.

[0003] Therefore, traditional aromatherapy mainly relies on the experience of perfumers, and consumers are usually unable to participate in the specific aromatherapy process. They can only choose ready-made aromatherapy or rely on the suggestions of perfumers.

[0004] Some businesses now allow consumers to customize aromatherapy on site, but since individual consumers lack experience in perfume making, it is difficult for them to make satisfactory aromatherapy on their own. Summary of the invention

[0005] In view of the above-mentioned defects, the purpose of the present invention is to propose a perfume-adjusting dual-arm robot system based on three-dimensional machine vision, which aims to intelligently generate multiple aromatherapy formulas that meet user needs, provide highly personalized aromatherapy modulation services, and meet the personalized needs of different users.

[0006] To achieve this object, the present invention adopts the following technical solutions:

[0007] A fragrance-adjusting dual-arm robot system based on three-dimensional machine vision, comprising the fragrance-adjusting dual-arm robot, wherein the fragrance-adjusting dual-arm robot is configured with a three-dimensional machine vision module, a fragrance-adjusting algorithm module and a mechanical arm equipped with a clamp;

[0008] The aromatherapy algorithm module is used to obtain the user's aromatherapy needs and generate a number of aromatherapy formulas for the user to choose based on the characteristics of the aromatherapy raw materials and the user's aromatherapy needs;

[0009] The three-dimensional machine vision module is used to identify and locate the positions of the required aromatherapy raw materials and aromatherapy containers after the user selects at least one aromatherapy formula, and to generate aromatherapy instructions for driving the mechanical arm and the clamping member;

[0010] The fragrance instruction is used to drive the robot arm and the clamp to perform weighing, adding, liquid transfer and stirring operations according to the required aromatherapy raw materials and the position of the fragrance container, the aromatherapy formula setting ratio and the fragrance order to obtain the prepared aromatherapy, bottle the prepared aromatherapy and pack it and output it to the user.

[0011] Preferably, the aromatherapy algorithm module is used to receive the aromatherapy requirements input by the user, generate a number of aromatherapy formulas, and visually output the aromatherapy formulas to the user for confirmation;

[0012] The aromatherapy requirements include fragrance level, concentration, scent type, usage occasion, season and durability.

[0013] Preferably, the three-dimensional machine vision module is used to pre-label the aromatherapy raw materials and the aromatherapy container, and after the user selects at least one aromatherapy formula, the module searches for the corresponding label according to the aromatherapy formula selected by the user to determine the location of the required aromatherapy raw materials and the aromatherapy container, and generates aromatherapy instructions for driving the mechanical arm and the clamping member;

[0014] After users provide feedback on the formula, incremental learning is performed based on their evaluation to optimize the generation of the next aromatherapy formula.

[0015] Furthermore, after receiving the fragrance adjustment instruction, the mechanical arm and the clamping member move the mechanical arm according to the fragrance adjustment sequence to drive the position of the clamping member to change, so that the clamping member moves to the position of the required aromatherapy raw material, and the clamping member clamps the required aromatherapy raw material according to the set ratio of the aromatherapy formula and weighs it, and then adds the weighed aromatherapy raw material into the fragrance adjustment container until all the required aromatherapy raw materials are added into the fragrance adjustment container according to the set ratio of the aromatherapy formula;

[0016] The clamping part clamps the stirring tool to stir the aroma raw materials in the aroma container to obtain the prepared aroma, and then adds the prepared aroma to different aroma bottles according to the net content of the aroma selected by the user, and reminds the user to collect it.

[0017] Preferably, in the aromatherapy algorithm module, generating a plurality of aromatherapy formulas based on the characteristics of aromatherapy raw materials and the aromatherapy needs of users includes:

[0018] Convert the user's aromatherapy needs into feature vectors;

[0019] Cluster analysis and decision tree are used to build user preference models, and user types are determined based on feature vectors;

[0020] Generate pre-selected aromatherapy formulas based on the characteristics and feature vectors of the raw materials, and use collaborative filtering solutions to generate several candidate aromatherapy formulas according to user types;

[0021] The pre-selected aromatherapy formula and the alternative aromatherapy formula are iteratively optimized to obtain several preferred aromatherapy formulas.

[0022] Furthermore, in the fragrance algorithm module, cluster analysis and decision tree are used to establish a user preference model, and the user type is determined based on the feature vector, including:

[0023] Use cluster analysis to group users with similar preferences into different types;

[0024] Analyze the feature mean of each user type to obtain the common preferences of each group;

[0025] Input the feature vector as user feature into the decision tree and perform importance analysis on the user feature;

[0026] Combining the results of user type and importance analysis, the aromatherapy formula trend is obtained.

[0027] Furthermore, in the aromatherapy algorithm module, generating a preselected aromatherapy formula according to the characteristics and feature vectors of the raw materials includes:

[0028] Obtain the characteristics of all raw materials;

[0029] Obtain raw material combinations based on the characteristics of the raw materials and the fragrance level required by the user;

[0030] Adjust the conventional mixing ratio according to the user's concentration requirements to obtain the latest concentration ratio;

[0031] The raw material combination and the latest concentration ratio are combined to obtain a pre-selected aromatherapy formula.

[0032] Furthermore, in the aromatherapy algorithm module, a collaborative filtering scheme is used to generate several candidate aromatherapy formulas according to user types, including:

[0033] Recommend aromatherapy formulas to users based on the formula choices of other users of the same user type and aromatherapy formula trends;

[0034] Combine the user type and use collaborative filtering algorithm to recommend aromatherapy formulas to obtain several first candidate aromatherapy formulas;

[0035] The first candidate aromatherapy formula is filtered and screened based on user needs to obtain a plurality of second candidate aromatherapy formulas.

[0036] Preferably, the fragrance adjustment instruction includes an aromatherapy combination code, and the aromatherapy combination code includes required aromatherapy raw materials and aromatherapy formula setting ratios.

[0037] One of the above technical solutions has the following advantages or beneficial effects:

[0038] The present invention is based on three-dimensional machine vision and can generate aromatherapy formulas according to the personalized needs of users. The three-dimensional machine vision module can identify the raw material position in real time, and the coordinated operation of the three-dimensional machine vision module and the robotic arm can be used to achieve high-precision raw material identification and preparation. The high degree of automation ensures that the robotic arm can accurately grab and put spices, avoid errors in manual perfume preparation, improve the accuracy of perfume preparation, and improve production efficiency and product quality. The perfume preparation algorithm module provides a variety of preferred formulas based on the characteristics of spices and user preferences to ensure the high consistency and high quality of aromatherapy products. It can monitor the inventory and usage of raw materials in real time, reduce the waste of raw materials caused by misoperation, reduce production costs, support users to choose different net aromatherapy contents and bottle styles, flexibly meet users' customized needs, and ensure the accuracy and quality of sample preparation. In addition, it can also be expanded to other fields that require precise proportions, such as cosmetics, soaps, etc., to achieve a wide range of cross-field applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0040] Figure 1 is a practical application diagram of a fragrance-adjusting dual-arm robot based on three-dimensional machine vision provided by an embodiment of the present invention;

[0041] Figure 2 is a flow chart of the fragrance blending of a fragrance blending dual-arm robot based on three-dimensional machine vision provided by an embodiment of the present invention;

[0042] Among them, there are a three-dimensional machine vision module 1, a robotic arm 2, and a clamping part 3. DETAILED DESCRIPTION

[0043] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0044] In the present invention, the terms "comprises", "comprising" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of more restrictions, an element defined by the sentence "comprising a ..." does not exclude the presence of other identical elements in the process, method, article or device comprising the element.

[0045] As consumers pursue personalized and high-quality life, the perfume industry needs to continuously launch innovative products that meet market demand. The preparation of traditional aromatherapy usually relies on the experience and manual operation of perfumers. The perfume making process includes the selection, proportioning and mixing of raw materials. The perfume making process often requires multiple tests and adjustments to ensure the concentration and fragrance of the aromatherapy. Due to its high dependence and labor intensity, the efficiency is low. Perfumers will constantly evaluate the fragrance of the aromatherapy and make adjustments as needed. This process requires repeated experiments to achieve the desired effect.

[0046] like Figure 1 As shown, a fragrance-adjusting dual-arm robot system based on three-dimensional machine vision includes the fragrance-adjusting dual-arm robot, which is configured with a three-dimensional machine vision module 1, a fragrance-adjusting algorithm module and a mechanical arm 2 equipped with a clamp 3;

[0047] The aromatherapy algorithm module is used to obtain the user's aromatherapy needs and generate a number of aromatherapy formulas for the user to choose based on the characteristics of the aromatherapy raw materials and the user's aromatherapy needs;

[0048] The three-dimensional machine vision module 1 is used to identify and locate the positions of the required aromatherapy raw materials and aromatherapy containers after the user selects at least one aromatherapy formula, and to generate aromatherapy instructions for driving the mechanical arm 2 and the clamping member 3;

[0049] The fragrance instruction is used to drive the robot arm 2 and the clamp 3 to perform weighing, adding, liquid transfer and stirring operations according to the required aromatherapy raw materials and the position of the fragrance container, the aromatherapy formula setting ratio and the fragrance order to obtain the prepared aromatherapy, and the prepared aromatherapy is bottled, packaged and output to the user.

[0050] The user's aromatherapy needs refer to the user's preferences when choosing aromatherapy, including but not limited to fragrance, concentration, durability, applicable occasions (such as daily use, relaxation, sleep, etc.), and seasonal preferences (such as fresh fragrance in summer and warm fragrance in winter).

[0051] There are many kinds of aromatherapy raw materials, each of which has its own unique aroma characteristics, volatility, concentration, duration and other characteristics. The extraction of raw material characteristics is the basis for preparing aromatherapy formulas. If the user chooses citrus fragrance as the "top note", a combination of spices including orange and lemon will be recommended. By referring to the feature vector, that is, referring to the user's needs, a preliminary aromatherapy formula can be obtained, which includes the raw materials of the formula and the corresponding concentrations.

[0052] Among them, each mechanical arm 2 of the perfume-adjusting dual-arm robot can perform different tasks. For example, one mechanical arm 2 is used to grab the raw materials and weigh them, and the other mechanical arm 2 can be responsible for stirring or transferring liquids; for another example, one mechanical arm 2 can be used to grab a perfume-adjusting container, such as a perfume-adjusting bottle, and the other mechanical arm 2 can grab an extraction tube, and use the extraction tube to extract the raw material fragrance liquid and drip it into the perfume-adjusting bottle. The movement of the mechanical arm 2 can be driven by a stepper motor or a servo motor, combined with a precise motion control algorithm to ensure the accuracy and coordination of each operation. The three-dimensional machine vision module 1 is the "eye" in the system. It captures images of the surrounding environment through devices such as cameras, laser radars, infrared sensors, etc., and analyzes the three-dimensional spatial position and shape of objects according to computer vision algorithms. The visual module can identify and locate perfume-adjusting raw materials and containers through deep learning algorithms. For example, a deep learning model (such as a convolutional neural network CNN) can classify and locate according to the characteristics of the object (such as color, shape, texture, etc.). The machine vision system can provide accurate position data for subsequent robot arm 2 operations by calculating the coordinates (XYZ coordinates) of the object in three-dimensional space.

[0053] The perfume algorithm module then analyzes the user's needs and generates multiple preferred perfume formulas based on the characteristics of the aromatherapy raw materials. For example, a raw material database can be established in advance. The database should contain information such as the name, characteristics, odor description, usage suggestions, interactions, etc. of each fragrance. Finally, several preferred formulas that meet the user's needs are generated. Each formula includes the raw materials and their usage ratios, which are ultimately provided for the user to choose.

[0054] The three-dimensional machine vision module 1 is responsible for accurately identifying the positions of aromatherapy raw materials and aromatherapy containers, and providing real-time visual feedback for the robot. In this embodiment, the aromatherapy dual-arm robot can be equipped with a fragrance adjustment area, and the fragrance adjustment area is configured with positions for placing adjusted aromatherapy products and various spices. After receiving the aromatherapy formula selected by the user, the three-dimensional machine vision module 1 identifies and locates the positions of the required aromatherapy raw materials and aromatherapy containers, and generates fragrance adjustment instructions for driving the robot arm 2 and the clamping member 3. When the robot arm 2 and the clamping member 3 receive the fragrance adjustment instructions, they calculate the required weight of each aromatherapy raw material according to the ratio set in the fragrance adjustment instructions, and the clamping member 3 clamps Take the aromatherapy raw materials and weigh them. After weighing the required weight, the robot arm 2 will extract the specified spices through the clamp 3 and transfer them to the aromatherapy container according to the adding ratio. After all the spices are added to the aromatherapy container, the left arm of the aromatherapy dual-arm robot can grab the pipette and the right arm can grab the test tube containing the spices. According to the preferred aromatherapy formula selected by the user, the spices in the required proportions are added to the test tubes in sequence. The prepared aromatherapy can be divided into several aromatherapy products according to different contents. For example, the bottling style is selected according to the net content of the aromatherapy selected by the user, and the prepared aromatherapy is automatically injected into the specified aromatherapy bottle. After the bottling is completed, the user is prompted to collect the aromatherapy.

[0055] The present invention is based on three-dimensional machine vision and can generate aromatherapy formulas according to the personalized needs of users. The three-dimensional machine vision module 1 recognizes the position of raw materials in real time, and utilizes the coordinated operation of the three-dimensional machine vision module 1 and the robot arm 2 to achieve high-precision raw material identification and preparation. The high degree of automation ensures that the robot arm 2 accurately grabs and puts spices, avoids errors in manual perfume preparation, improves the accuracy of perfume preparation, and improves production efficiency and product quality. The formula generation module provides a variety of preferred formulas based on the characteristics of spices and user preferences to ensure the high consistency and high quality of aromatherapy products. It can monitor the inventory and usage of raw materials in real time, reduce the waste of raw materials caused by misoperation, reduce production costs, support users to choose different net aroma content and bottle styles, flexibly meet the customized needs of users, and ensure the accuracy and quality of sample preparation. In addition, Figure 1 The perfume-making dual-arm robot shown is not only suitable for aromatherapy preparation, but can also be extended to other fields that require precise proportions, such as cosmetics, soaps, etc., to achieve a wide range of cross-field applications.

[0056] Preferably, the aromatherapy algorithm module is used to receive the aromatherapy requirements input by the user, generate a number of aromatherapy formulas, and visually output the aromatherapy formulas to the user for confirmation;

[0057] The aromatherapy requirements include fragrance level, concentration, scent type, usage occasion, season and durability.

[0058] Fragrance is the main component of aromatherapy, which is usually divided into three layers: top note (initial aroma), middle note (the core of the aroma) and base note (the lasting aftertaste of the aroma). Users can specify what kind of combination of fragrance levels they want the aromatherapy to have, such as floral, fruity, woody, vanilla, etc. Concentration refers to the intensity of the aromatherapy or the intensity of the fragrance. Users can choose different concentrations of aromatherapy, such as "fresh" and "strong", which will affect the proportion of different spices used in the aromatherapy formula. Users can choose the classification of smells, such as floral, fruity, woody, vanilla, citrus, oriental, etc. Different types of smells represent different combinations and blending methods of spices. The occasion of use of aromatherapy determines the selection and matching of spices in the formula. For example, aromatherapy for home use may tend to be warm and soothing, while aromatherapy for offices or public places may choose fresh and refreshing spices. The seasonal requirements of aromatherapy are also an important factor in the process of perfume making. Winter aromatherapy tends to be warm and calm (such as woody and spice scents), while summer aromatherapy tends to be fresh and light (such as citrus and floral scents). The durability of aromatherapy refers to the time the aroma lasts in the air. Users can choose "long-lasting" or "fast evaporation", which will affect the ratio of volatile components to persistent components in the aromatherapy formula.

[0059] In this embodiment, the scenting algorithm module will analyze the aromatherapy needs input by the user and map these needs to the available spice database. Each spice has different properties such as smell, volatility, concentration, and persistence. The algorithm will match according to these properties. For example, floral fragrance is suitable for top notes, while woody fragrance is usually used for base notes. According to the different needs of users, the algorithm will assign a certain weight to each requirement (fragrance level, concentration, odor type, season, etc.). If the user particularly emphasizes "persistence", the algorithm will add ingredients with longer lasting fragrance time (such as wood and vanilla spices) to the aromatherapy formula. After generating several aromatherapy formulas, the scenting dual-arm robot system will visualize the aromatherapy formula to the user through a graphical interface or an interactive interface, such as showing the position and proportion of different spices in the top notes, middle notes, and base notes. The user can see the name, concentration, and volatility characteristics of each layer of spices. Through labels or color coding, the user is intuitively shown the classification of aromas, such as floral, fruity, woody, etc., and the distribution of aroma concentration and persistence is displayed to help the user intuitively understand the aroma intensity and persistence of each formula. Users can adjust certain parameters of the formula (such as increasing or decreasing the proportion of a certain fragrance) through the interface, and view the changes in the visual results in real time, and finally confirm the most suitable aromatherapy formula. Once the user selects a final formula, the aromatherapy algorithm module will output the formula as specific formula data and generate aromatherapy instructions for subsequent operations.

[0060] Preferably, the three-dimensional machine vision module 1 is used to pre-label aromatherapy raw materials and fragrance containers. After the user selects at least one aromatherapy formula, the corresponding label is searched according to the aromatherapy formula selected by the user to determine the location of the required aromatherapy raw materials and fragrance containers, and the fragrance instructions for driving the robot arm 2 and the clamp 3 are generated; after the user provides feedback on the formula, incremental learning is performed based on the user's evaluation to optimize the generation of the next aromatherapy formula.

[0061] Each piece of aromatherapy raw material and container will be affixed with a label with a unique identifier (such as a QR code, barcode or visual label). After the user selects the aromatherapy formula, the system will identify the required label based on the aromatherapy raw material and container information in the formula. The three-dimensional machine vision module 1 identifies the labels of the aromatherapy raw materials and containers and locates their specific positions in the workspace. By identifying the labels, the system can accurately know the current positions of the aromatherapy raw materials and containers, and then transmit this position information to the system. According to the identified label information, the action that the robot arm 2 needs to perform is determined, for example, which bottle of aromatherapy raw material or which container the gripper needs to grab, and the best path for the robot arm 2 to reach the target raw material or container is calculated, taking into account factors such as obstacles in the space and the placement of raw materials.

[0062] Further, after receiving the fragrance adjustment instruction, the robot arm 2 and the clamping member 3 move the robot arm 2 according to the fragrance adjustment sequence to drive the position of the clamping member 3 to change, so that the clamping member 3 moves to the position of the required aromatherapy raw material, and the clamping member 3 clamps the required aromatherapy raw material according to the set ratio of the aromatherapy formula, weighs it, and then adds the weighed aromatherapy raw material into the fragrance adjustment container until all the required aromatherapy raw materials are added into the fragrance adjustment container according to the set ratio of the aromatherapy formula;

[0063] The clamping member 3 clamps the stirring tool to stir the aroma raw materials in the aroma container to obtain the prepared aroma, and then adds the prepared aroma to different aroma bottles according to the net content of the aroma selected by the user, and reminds the user to collect it.

[0064] When the robot arm 2 and the clamp 3 receive the fragrance instruction, according to the order set in the fragrance formula, the robot arm 2 is controlled to move and drive the clamp 3 to adjust to the position of the required aromatherapy raw materials. The clamp 3 accurately clamps the required aromatherapy raw materials according to the set ratio in the formula, and weighs them to ensure the accurate amount of each raw material. After weighing, the clamp 3 adds the aromatherapy raw materials to the fragrance container one by one until all the raw materials are added according to the aromatherapy formula ratio. Then, the clamp 3 takes out the stirring tool and fully stirs the aromatherapy raw materials in the container until the aromatherapy that meets the requirements is prepared. Finally, the system evenly distributes the prepared aromatherapy to different aromatherapy bottles according to the net content of the aromatherapy selected by the user, and reminds the user to pick up the prepared aromatherapy products after completion. The whole process is efficient and accurate, ensuring the quality of each bottle of aromatherapy and the personalized needs of the user.

[0065] In the aromatherapy algorithm module, several aromatherapy formulas are generated based on the characteristics of aromatherapy raw materials and the aromatherapy needs of users, including:

[0066] Convert the user's aromatherapy needs into feature vectors;

[0067] Cluster analysis and decision tree are used to build user preference models, and user types are determined based on feature vectors;

[0068] Generate pre-selected aromatherapy formulas based on the characteristics and feature vectors of the raw materials, and use collaborative filtering solutions to generate several candidate aromatherapy formulas according to user types;

[0069] The pre-selected aromatherapy formula and the alternative aromatherapy formula are iteratively optimized to obtain several preferred aromatherapy formulas.

[0070] The aromatherapy needs input by users are often unstructured and difficult to process directly with traditional methods. Therefore, they need to be converted into a form that can be understood by the machine through some methods, so they are converted into feature vectors. Through feature vectors, we can convert user demand information into digital format for further analysis and calculation. Each feature corresponds to a different dimension of user preference, for example: Fragrance level: top note: citrus = 1, middle note: floral = 2, base note: woody = 3; concentration: low = 1, medium = 2, high = 3; smell type: fresh and natural = 1 (simplified, map this description to a numerical value); usage occasions: home = 1, office = 2, outdoor = 3; seasons: spring = 1, summer = 2, autumn = 3, winter = 4; duration: short time = 1, medium time = 2, long time = 3. Assuming the user input is top note: citrus (such as lemon), middle note: floral (such as rose), base note: woody (such as sandalwood); concentration: medium; smell type: hope it is a fresh and natural combination; occasion of use: home, especially living room; season: spring; persistence: hope the fragrance can last for a long time, it is converted into a feature vector: [1, 2, 3, 2, 1, 1, 1, 3]. Cluster analysis is a method of unsupervised learning that aims to group data according to their similarities. Cluster analysis calculates the similarity between data points (for example, using Euclidean distance, cosine similarity, etc.) to classify users with similar needs into one category. Cluster analysis determines the similarity between data points by selecting appropriate metrics. For example, the Euclidean distance between the feature vectors of user needs is calculated. If the distance is small, it means that the needs of the two users are similar and can be classified into the same category. The commonly used algorithms in cluster analysis are K-means and DBSCAN. These algorithms automatically discover the inherent structure in user needs and then divide users into different groups. Through cluster analysis, user groups with similar needs can be identified, and similar aromatherapy formulas can be recommended to these users. Decision tree is a classification and regression method in supervised learning. It classifies users based on features. Its working principle is to fork according to a series of conditions (features) until it reaches a leaf node. The leaf node represents the predicted result. The decision tree selects the feature that can maximize the "information gain" or "Gini coefficient" to split the node and gradually assign users to different categories. Information gain is a measure of the influence of the feature on the classification result. A larger information gain indicates that the feature has a greater effect on the data division. In aromatherapy blending, the decision tree model can predict the type of aromatherapy that the user is most likely to like based on the user's characteristics (such as fragrance, concentration preference, etc.). Through cluster analysis and decision trees, we can assign a "type" to each user. These types represent the user's preference pattern in aromatherapy selection, such as: "Users who prefer a higher concentration of floral fragrance", "Users who like fresh fragrance", etc. The determination of user type helps the subsequent recommendation system to provide users with customized aromatherapy formulas.Collaborative filtering is a recommendation algorithm based on user historical behavior or similar user group behavior. Its basic principle is: if user A and user B have similar preferences, then the aromatherapy formulas they like are also likely to be similar. Therefore, the previous step determines the type of this user. In this step, it is necessary to analyze the formulas selected by users of the same type. If user A and user B have similarities in their choices, and user B chooses a specific formula, it will be considered to recommend it to user A. Therefore, in the process of aromatherapy preparation, collaborative filtering recommends multiple possible aromatherapy formulas for each user according to the types (i.e., needs) of different users. Iterative optimization of pre-selected aromatherapy formulas and alternative aromatherapy formulas can be obtained by genetic algorithm processing, including selecting an initial population: selecting pre-selected aromatherapy formulas and alternative aromatherapy formulas as the initial population; fitness evaluation: calculating the fitness of each formula in the initial population; crossover mutation: selecting a formula with high fitness as the parent generation, and performing crossover to generate the first new formula, performing random mutation on the first new formula to obtain the second new formula; repeating fitness evaluation and crossover mutation until the preset number of iterations is reached, and obtaining several preferred aromatherapy formulas and their instructions for use. First, a certain number of formulas are selected from the pre-selected aromatherapy formulas and the alternative aromatherapy formulas as the initial population. Each aromatherapy formula can be regarded as a chromosome, which contains information such as the type, proportion and smell type of the fragrance. For example, an aromatherapy formula may contain lavender, sandalwood and citrus, and the proportion of the fragrance is 50%: 30%: 20%. Then the fitness of each formula is evaluated. The fitness can reflect the quality of the formula. The fitness function can be based on the following indicators: fragrance level, persistence, usage occasion and concentration. After the fitness evaluation, the aromatherapy formula with higher fitness is selected as the parent generation to generate the next generation of formulas. Common selection methods include roulette selection and tournament selection. The higher the fitness value, the greater the probability of the formula being selected. For example, if the fitness of a formula is 90%, it has a higher probability of being selected as the parent generation. Two aromatherapy formulas with higher fitness are selected for crossover to simulate the process of gene recombination. The crossover method can be to exchange some components of the parent generation to generate a new formula.For example, the crossover of formula A and formula B: new formula: top note 30%, middle note 50%, base note 20%. The crossover operation can be performed in multiple aspects such as the type, proportion, and aroma characteristics of the spices. The crossover operation will increase the diversity of the formula. A certain random mutation will be performed on the basis of the generated new formula. The mutation can be achieved by changing the proportion of spices, adding or removing certain spices. For example, the second new formula of the new formula after mutation is 31% top note, 49% middle note, and 20% base note. The second new formula will then be iterated, and the fitness evaluation and crossover mutation operations will be repeated for multiple rounds of iterations. A new formula will be generated in each generation, and the fitness, selection, crossover and mutation will be continuously evaluated. The number of iterations is usually preset, such as 50 or 100 generations, or until a preset threshold is reached. After multiple rounds of iterations, the genetic algorithm will output several preferred aromatherapy formulas. Each preferred aromatherapy formula will be accompanied by corresponding instructions for use, including the specific ingredients of the formula, suitable usage occasions, aroma intensity, persistence, etc., to help users better choose and use aromatherapy.

[0071] The calculation of fitness satisfies the relationship:

[0072] Fitness=w1*P top *S top +w2*P middle *S middle +w3*P base *S base +w4*B score +w5*D score ;

[0073] Among them, w1 represents the weight coefficient of the top note, w2 represents the weight coefficient of the middle note, w3 represents the weight coefficient of the base note, w4 represents the weight coefficient of balance, w5 represents the weight coefficient of persistence, P top Indicates the pre-adjustment ratio, P middle Indicates the mid-adjustment ratio, P base Indicates the post-adjustment ratio, S top Indicates the top note score, S middle Indicates a medium-tone rating, S base Indicates the post-adjustment score, B score Indicates the balance score, D score Represents the persistence score.

[0074] Furthermore, in the fragrance algorithm module, cluster analysis and decision tree are used to establish a user preference model, and the user type is determined based on the feature vector, including:

[0075] Use cluster analysis to group users with similar preferences into different types;

[0076] Analyze the feature mean of each user type to obtain the common preferences of each group;

[0077] Input the feature vector as user feature into the decision tree and perform importance analysis on the user feature;

[0078] Combining the results of user type and importance analysis, the aromatherapy formula trend is obtained.

[0079] Specifically, in another embodiment, cluster analysis can use the K-means algorithm, which is an unsupervised learning clustering algorithm that divides the same user type into the same cluster. By minimizing the differences in samples within the cluster, the data within each cluster is made as similar as possible, while the data differences between different clusters are as large as possible, so as to understand the common fragrance preferences of different user groups. For example, in K-means clustering, the "features" of fragrance preference refer to the fragrance (floral, woody, fruity, etc.), concentration (light fragrance, strong fragrance, etc.), and usage occasions (daytime, night, special occasions, etc.). The fragrance preference of each user can be represented by these features, and the Euclidean distance is used to measure the similarity between data points. In the fragrance preference analysis, the closer the distance between users, the more similar their fragrance preferences are. Based on this principle, the needs of each user are calculated, and the user types can be distinguished. For example, type A prefers fresh fragrance or warm woody fragrance.

[0080] By inputting the feature vector as a user feature into the decision tree, the decision tree can automatically consider a certain feature (such as fragrance, concentration, occasion of use, etc.) as the most important for the user's fragrance selection. By calculating the importance of the feature (such as the Gini index or information gain), it can be determined which fragrance features can best drive the user's choice. Based on the user's fragrance features, the decision tree can be used to predict the user's fragrance preference, or classify the user, such as classifying the user into groups that "prefer fresh fragrance" or "prefer warm fragrance", and determine which features can best predict the user's perfume choice, thereby revealing potential fragrance preference patterns. The results of preference and importance analysis of different user groups are combined to identify current market trends. For example, if multiple groups show a strong preference for a certain fragrance, it means that the formula of this fragrance may be becoming a popular trend in the market.

[0081] Furthermore, in the aromatherapy algorithm module, generating a preselected aromatherapy formula according to the characteristics and feature vectors of the raw materials includes:

[0082] Obtain the characteristics of all raw materials;

[0083] Obtain raw material combinations based on the characteristics of the raw materials and the fragrance level required by the user;

[0084] Adjust the conventional mixing ratio according to the user's concentration requirements to obtain the latest concentration ratio;

[0085] The raw material combination and the latest concentration ratio are combined to obtain a pre-selected aromatherapy formula.

[0086] Before executing the present modulation method, a database can be constructed, which contains information such as the fragrance, odor characteristics, suitable scenarios and seasonal usage recommendations of each raw material. When necessary, the characteristics of the raw materials can be obtained from the database to facilitate subsequent aromatherapy modulation.

[0087] According to the basic theory of perfume blending, the fragrance of perfume is divided into three levels: top notes, middle notes and base notes. The fragrances at each level have different volatility characteristics and odor intensity. Therefore, it is necessary to screen out the appropriate fragrance combination from the fragrance database according to the user's fragrance level preference. Among them, the top notes have a faster volatilization speed and are usually the first smell to be perceived. The top notes usually choose fragrances with strong volatility and fresh smell, such as citrus (orange, lemon) or fresh mint and herbal plants. The middle note is also called the heart note. The smell appears a few minutes to a few hours after the perfume is sprayed. It is usually softer and has a certain sense of hierarchy. Common fragrances include floral fragrance (such as rose, jasmine) and vanilla and herbs. The base note is the base tone of the perfume. The smell is the most lasting. It is usually a warmer and deeper fragrance, such as woody fragrance (sandalwood, cedar), spices (such as vanilla, amber), etc. For example, if the user chooses the "top note" preference as citrus fragrance, a combination of spices including orange, lemon, etc. will be recommended.

[0088] The concentration ratio in the aromatherapy formula determines the balance of the spices in each fragrance level. The common blending ratio is: 30% top note, 50% middle note and 20% base note, but different raw material types and user needs may require different concentration ratios. After the raw material combination is determined in the previous step, the most appropriate raw material ratio is given according to user needs to ensure that the blended aromatherapy achieves the best balance in aroma. The raw material combination and concentration ratio are combined to obtain the pre-selected aromatherapy formula.

[0089] Furthermore, in the aromatherapy algorithm module, a collaborative filtering scheme is used to generate several candidate aromatherapy formulas according to user types, including:

[0090] Recommend aromatherapy formulas to users based on the formula choices of other users of the same user type and aromatherapy formula trends;

[0091] Combine the user type and use collaborative filtering algorithm to recommend aromatherapy formulas to obtain several first candidate aromatherapy formulas;

[0092] The first candidate aromatherapy formula is filtered and screened based on user needs to obtain a plurality of second candidate aromatherapy formulas.

[0093] Aromatherapy formula trends can reflect current market trends, and the formulas of other users of the same user type can represent potential aromatherapy formulas corresponding to the current user's needs. Aromatherapy formula trends can be used to screen the formulas of other users of the same user type and recommend them to users. The behavior of similar user groups can be used to filter potential aromatherapy choices, ensuring that the recommendation system not only considers the preferences of the current user, but also the group trends of user types, greatly improving the accuracy and relevance of recommendations.

[0094] Using collaborative filtering algorithms, multiple alternative formulas corresponding to the recommended formulas will be generated based on the preferences of similar users. The algorithm will give priority to user groups with similar characteristics when finding similar users. For example: Formula 1: Top note: 30% lemon, middle note: 50% rose, base note: 20% amber; Formula 2: Top note: 25% mint, middle note: 55% ocean, base note: 20% vanilla; Formula 3: Top note: 35% orange, middle note: 45% lavender, base note: 20% sandalwood, as the first alternative aromatherapy formula, but these formulas may only be selected based on preference similarity and user type, and may not necessarily meet the specific needs of each user. Therefore, it is necessary to screen the first alternative scheme in combination with user needs, eliminate the formulas that do not meet the needs, and only retain the aromatherapy formulas that meet the needs to obtain the second alternative aromatherapy formula. For example, if a user prefers a strong woody scent and is a business user, then even if collaborative filtering recommends a fresh floral aromatherapy formula, this aromatherapy formula will be screened out because it does not meet the user's needs. If the user needs a perfume suitable for summer, the formula with fresh fragrance is recommended first. Through collaborative filtering algorithms and user type analysis, accurate recommendations can be made based on the user's historical behavior and the preferences of similar users. When user needs are relatively rare, collaborative filtering can make effective recommendations based on the behavior of similar users, which can alleviate the common data sparsity problem in recommendation operations.

[0095] Preferably, the fragrance adjustment instruction includes an aromatherapy combination code, and the aromatherapy combination code includes required aromatherapy raw materials and aromatherapy formula setting ratios.

[0096] Specifically, an aromatherapy combination code is also proposed, which assigns a unique number to each fragrance to form a raw material database. The database should contain information such as the name, characteristics, odor description, usage suggestions and interactions of each fragrance. The raw materials can be classified by type (plant extracts, synthetic fragrances, natural fragrances, etc.), and labels (such as top notes, middle notes, and base notes) are added to facilitate the selection of subsequent fragrances. When the user selects a group of raw materials for fragrance, a combination code is generated according to the number and amount of each fragrance. The combination code not only indicates the type of fragrance used, but also reflects the proportional relationship of each fragrance in the formula. For example, the raw material number and its corresponding amount are used to form a format similar to "raw material number. Amount" for quick identification and analysis. Example: If three spices are selected, the code is "001.30-002.20-003.50", which means that the amount of fragrance numbered 001 is 30%, the amount of fragrance numbered 002 is 20%, and the amount of fragrance numbered 003 is 50%, which can be more convenient for robot recognition.

[0097] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "illustrative embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0098] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.

Claims

1. A fragrance-adjusting dual-arm robot system based on three-dimensional machine vision, characterized in that: The dual-arm fragrance-adjusting robot comprises a three-dimensional machine vision module, a fragrance-adjusting algorithm module and a mechanical arm equipped with a clamp; The aromatherapy algorithm module is used to obtain the user's aromatherapy needs and generate a number of aromatherapy formulas for the user to choose based on the characteristics of the aromatherapy raw materials and the user's aromatherapy needs; The three-dimensional machine vision module is used to identify and locate the positions of the required aromatherapy raw materials and aromatherapy containers after the user selects at least one aromatherapy formula, and to generate aromatherapy instructions for driving the mechanical arm and the clamping member; The fragrance instruction is used to drive the robot arm and the clamp to perform weighing, adding, liquid transfer and stirring operations according to the required aromatherapy raw materials and the position of the fragrance container, the aromatherapy formula setting ratio and the fragrance order to obtain the prepared aromatherapy, bottle the prepared aromatherapy and pack it and output it to the user.

2. The fragrance blending dual-arm robot system according to claim 1, characterized in that: The aromatherapy algorithm module is used to receive the aromatherapy requirements input by the user, generate a number of aromatherapy formulas, and visually output the aromatherapy formulas to the user for confirmation; The aromatherapy requirements include fragrance level, concentration, scent type, usage occasion, season and durability.

3. The fragrance blending dual-arm robot system according to claim 1, characterized in that: The three-dimensional machine vision module is used to pre-label the aromatherapy raw materials and the aromatherapy container. After the user selects at least one aromatherapy formula, the module searches for the corresponding label according to the aromatherapy formula selected by the user to determine the location of the required aromatherapy raw materials and the aromatherapy container, and generates aromatherapy instructions for driving the robot arm and the clamping member; After users provide feedback on the formula, incremental learning is performed based on their evaluation to optimize the generation of the next aromatherapy formula.

4. The fragrance blending dual-arm robot system according to claim 3, characterized in that: After receiving the fragrance adjustment instruction, the mechanical arm and the clamping part move the mechanical arm according to the fragrance adjustment sequence to drive the position of the clamping part to change, so that the clamping part moves to the position of the required aromatherapy raw material, and the clamping part clamps the required aromatherapy raw material according to the set ratio of the aromatherapy formula and weighs it, and then adds the weighed aromatherapy raw material into the fragrance adjustment container until all the required aromatherapy raw materials are added into the fragrance adjustment container according to the set ratio of the aromatherapy formula; The clamping part clamps the stirring tool to stir the aroma raw materials in the aroma container to obtain the prepared aroma, and then adds the prepared aroma to different aroma bottles according to the net content of the aroma selected by the user, and reminds the user to collect it.

5. The fragrance blending dual-arm robot system according to claim 1, characterized in that: In the aromatherapy algorithm module, several aromatherapy formulas are generated based on the characteristics of aromatherapy raw materials and the aromatherapy needs of users, including: Convert the user's aromatherapy needs into feature vectors; Cluster analysis and decision tree are used to build user preference models, and user types are determined based on feature vectors; Generate pre-selected aromatherapy formulas based on the characteristics and feature vectors of the raw materials, and use collaborative filtering solutions to generate several candidate aromatherapy formulas according to user types; The pre-selected aromatherapy formula and the alternative aromatherapy formula are iteratively optimized to obtain several preferred aromatherapy formulas.

6. The fragrance blending dual-arm robot system according to claim 5, characterized in that: In the fragrance algorithm module, cluster analysis and decision tree are used to establish a user preference model, and the user type is determined based on the feature vector, including: Use cluster analysis to group users with similar preferences into different types; Analyze the feature mean of each user type to obtain the common preferences of each group; Input the feature vector as user feature into the decision tree and perform importance analysis on the user feature; Combining the results of user type and importance analysis, the aromatherapy formula trend is obtained.

7. The fragrance blending dual-arm robot system according to claim 6, characterized in that: In the aromatherapy algorithm module, generating a preselected aromatherapy formula according to the characteristics and feature vectors of the raw materials includes: Obtain the characteristics of all raw materials; Obtain raw material combinations based on the characteristics of the raw materials and the fragrance level required by the user; Adjust the conventional mixing ratio according to the user's concentration requirements to obtain the latest concentration ratio; The raw material combination and the latest concentration ratio are combined to obtain a pre-selected aromatherapy formula.

8. The fragrance blending dual-arm robot system according to claim 7, characterized in that: In the aromatherapy algorithm module, a collaborative filtering scheme is used to generate several candidate aromatherapy formulas according to user types, including: Recommend aromatherapy formulas to users based on the formula choices of other users of the same user type and aromatherapy formula trends; Combine the user type and use collaborative filtering algorithm to recommend aromatherapy formulas to obtain several first candidate aromatherapy formulas; The first candidate aromatherapy formula is filtered and screened based on user needs to obtain a plurality of second candidate aromatherapy formulas.

9. The fragrance blending dual-arm robot system according to claim 1, characterized in that: The fragrance adjustment instruction includes an aromatherapy combination code, and the aromatherapy combination code includes required aromatherapy raw materials and aromatherapy formula setting ratios.

Images